A source network load storage collaborative interaction optimization control method

Abstract

The application discloses a kind of source network load storage collaborative interaction optimization control method, solve power grid and distributed resource two-way regulation, communication delay and cross-regional coordination problem, belong to energy power technology field.In the method, first, the multi-objective optimization model of integrating distributed energy output, flexible load, energy storage charging and discharging and power grid state is constructed, to generate initial collaborative scheduling scheme;Determine resource allocation ratio by full-link iterative calculation, combined with delay compensation mechanism Dynamic adjustment signal threshold and edge response plan;Fusion autonomous response mechanism and equipment communication protocol extension field, generate protocol supported autonomous response plan;Through wide-area resource coordination parameter correction, realize local to wide-area dynamic adjustment;Finally, aggregate and evaluate voltage stability and power balance degree, output optimization result to each link resource, improve system coordination efficiency and stability.

Description

Technical Field

[0001] This invention relates to the field of energy and power technology, and in particular to a method for coordinated and interactive optimization control of power generation, grid, load and storage. Background Technology

[0002] With the accelerated transformation of the global energy structure, the penetration rate of distributed energy sources, represented by photovoltaics and wind power, in the power grid continues to increase, driving the power system to evolve from the traditional "source follows load" model to a collaborative interaction model of "source-grid-load-storage". In the context of building a new power system, how to achieve real-time scheduling and wide-area optimization of distributed energy sources through edge computing and cloud collaboration has become a core technological challenge in the field of the energy internet.

[0003] In existing technologies, two main approaches are used for the coordinated scheduling of distributed energy resources and the power grid: one is a cloud-based centralized optimization scheduling mode that builds a global optimization model by collecting data from the entire network; the other is an edge device autonomous response mechanism that relies on local real-time data (such as energy storage SOC and load curves) for local control.

[0004] However, neither of these two approaches effectively resolves the deep-seated coupling contradiction between "local real-time response" and "wide-area consistent control." Specifically, the fixed-cycle communication mechanism at the edge side is mismatched with the dynamic time delay distribution across regions, leading to excessive delays in control signals and amplified real-time feedback deviations. Furthermore, when cloud-based optimization schemes are distributed locally, the lack of efficient delay compensation and parameter correction mechanisms easily causes fluctuations in voltage stability and power balance indicators, and generates a time delay accumulation effect in cross-regional collaboration. Ultimately, this results in a control gap during the transition from edge plans to wide-area protocols, making it difficult for the overall energy interaction effect to reach the theoretical optimal level. Summary of the Invention

[0005] This invention aims to provide a source-grid-load-storage collaborative interaction optimization control method, which realizes "spatiotemporal adaptation, two-way interaction, delay perception, protocol fusion, and closed-loop optimization" of source-grid-load-storage collaborative interaction, thereby improving the energy utilization efficiency, power supply reliability, and control accuracy of the power system.

[0006] The technical solution adopted in this invention is:

[0007] A source-grid-load-storage coordinated interactive optimization control method includes the following steps:

[0008] Step S1: Construct a multi-objective optimization model, integrate relevant parameters of distributed energy output, flexible load regulation capability, energy storage system charging and discharging strategy and overall grid operation status to obtain an initial collaborative scheduling scheme;

[0009] Step S2: Based on the initial coordinated scheduling scheme, iterative calculations are performed through the entire energy interaction process to determine the final resource allocation ratio under the two-way mutual adjustment mechanism between the power grid and distributed energy.

[0010] Step S3: Generate control signals based on the final resource allocation ratio, and obtain an edge distribution response plan containing the control release signal sequence through real-time feedback from local devices and global load balancing optimization of edge nodes;

[0011] Step S4: Evaluate the communication delay level and real-time feedback response deviation of the control and release signal sequence, integrate the delay compensation mechanism, and determine the adjustment threshold of the delay-sensitive signal;

[0012] Step S5: Integrate the delay-sensitive signal adjustment threshold into the correction process of the wide-area resource collaboration parameters to obtain the wide-area resource interaction correction parameters and signal release adjustment scheme;

[0013] Step S6: Based on the wide area resource interaction correction parameters and signal release adjustment scheme, integrate the autonomous adjustment parameter update path of the autonomous response mechanism with the message triggering and response interface of the device communication protocol, and determine the response coordination scheme under the embedded protocol extension field configuration.

[0014] Step S7: Use a response coordination scheme to process the local resource allocation execution unit in the edge distribution response plan, generate an adjustment sequence of the integrated device communication protocol delay management rule set, and obtain the autonomous response plan supported by the protocol.

[0015] Step S8: Combine the wide-area resource interaction correction parameters and generate the final interaction scheme based on the autonomous response plan;

[0016] Step S9: Obtain the status index data in the final interactive solution, aggregate and evaluate the overall effect of distributed energy output and energy storage system charging and discharging, obtain the final optimized output, and execute signal distribution to resources in each link.

[0017] Furthermore, in step S1, the dimensional expansion of time and space attributes is incorporated to optimize the initial collaborative scheduling scheme.

[0018] Furthermore, the process of step S2 includes:

[0019] From the initial coordinated scheduling scheme, analyze and extract energy interaction data and initial resource allocation ratios for each stage;

[0020] Perform dynamic analysis of energy interaction across the entire process, calculate the interaction intensity and regulation demand in layers, and evaluate the grid regulation capacity and distributed energy response capacity in conjunction with the two-way mutual regulation mechanism between the grid and distributed energy.

[0021] The initial resource allocation ratio is iteratively processed, and the grid regulation capacity and distributed energy response capacity are balanced through algorithm optimization. An optimized allocation scheme that meets the requirements of energy interaction balance in the whole process is output, verified, and the final resource allocation ratio is output.

[0022] Furthermore, the process of step S3 includes:

[0023] Extract core variables from the final resource allocation ratio, assign weights, and generate control signals through time series analysis algorithms;

[0024] The control signal is transmitted to the local device to perform status judgment and frequency adjustment, and then the local device is driven to perform real-time feedback and return real-time feedback data.

[0025] Based on real-time feedback data, the correspondence between signal strength and resource adjustment amount is analyzed to generate autonomous response execution rules and output response mapping paths;

[0026] Based on the response mapping path, the load balancing performance of the edge distribution scheme is evaluated and verified, and then a final edge distribution response plan containing the control and release signal sequence is generated. Further, step S4 includes the following steps:

[0027] For the control and release of signal sequences, collect signal transmission data, and statistically analyze delay data and delay level classification;

[0028] The preset delay compensation mechanism is invoked based on the delay level classification to adjust the signal delivery sequence;

[0029] Collect real-time feedback response data, perform deviation threshold judgment, and optimize the feedback cycle;

[0030] Key time nodes are extracted from the optimized feedback cycle. Combined with the delay-sensitive characteristics, a preliminary signal is output to adjust the threshold range and calibrate. Finally, a delay-sensitive signal is output to adjust the threshold.

[0031] Furthermore, the process of step S5 includes:

[0032] The initial threshold data for signal adjustment is extracted and analyzed from the delay-sensitive signal adjustment threshold.

[0033] The initial threshold data is compared with the preset standard, and combined with the cross-regional time delay distribution, to obtain the response deviation value or response deviation evaluation data. The response deviation value is the instantaneous feedback deviation calculated based on the cross-regional time delay mean, and the response deviation evaluation data is the response deviation threshold after cross-regional time delay distribution analysis and expansion.

[0034] Based on the allocation of wide-area resources and the needs of business scenarios, the response deviation value or response deviation evaluation data is corrected to obtain the wide-area resource interaction correction parameters.

[0035] Based on the wide-area resource interaction correction parameters, the processing flow for adjusting the threshold of delay-sensitive signals and the cross-regional signal synchronization strategy are optimized to obtain a signal release adjustment scheme.

[0036] Furthermore, the process of step S6 is as follows:

[0037] Acquire the communication interface data stream of the edge device and classify the message trigger time points and content contained therein;

[0038] Based on the message trigger classification results, the extended field structure of the device communication protocol is parsed, and field-trigger mapping and priority sequence judgment are performed;

[0039] Based on the priority sequence configured by the field, the adjustment logic of the autonomous response mechanism and the signal release adjustment scheme are integrated to output the initial scheme for path adjustment;

[0040] The path adjustment initial scheme is used to adjust the interaction configuration of the device communication interface in real time, and the actual interaction data of the device communication interface is collected, compared with the interaction configuration, and the execution strategy is output.

[0041] Update the interaction configuration of the device communication interface, synchronously write the updated configuration into the protocol extension field, verify it, and output the response coordination scheme embedded in the protocol extension field configuration.

[0042] Furthermore, the process of step S7 is as follows:

[0043] Acquire real-time status data of edge distributed nodes, determine the priority ranking of key nodes, obtain local resource availability information based on the ranking results, and output a preliminary resource allocation plan.

[0044] For the initial resource allocation plan, collect the interaction logs of edge devices, integrate the latency management history records, and output a set of rules for latency optimization;

[0045] The edge device execution is driven by a delay optimization rule set, the execution feedback is collected and analyzed, and a response mechanism configuration scheme is output.

[0046] Based on the response mechanism configuration scheme, obtain the compatibility data supported by the protocol, and generate an autonomous response plan supported by the protocol after verification.

[0047] Furthermore, the process of step S8 is as follows:

[0048] Calculate the deviation between the autonomous response plan and the cloud optimization paradigm, analyze the source of the deviation or verify the parameter benchmark, use optimization algorithms to analyze the wide-area resource interaction correction parameters, and output a set of wide-area resource collaboration parameters;

[0049] By combining the cross-regional energy allocation index in the wide-area resource interaction correction parameters, the corresponding resource coordination parameters in the wide-area resource coordination parameter set are further decomposed, and regional energy allocation schemes are output and verified.

[0050] By combining regional energy allocation schemes with wide-area resource interaction correction parameters and optimizing them, the final interaction scheme is output.

[0051] Furthermore, the process of step S9 is as follows:

[0052] State index data is extracted from the final interactive solution, classified, and then a set of classification indicators is output; among them, the state index data includes voltage stability and power balance.

[0053] The voltage stability and power balance are evaluated; if the preliminary evaluation results show that a certain indicator exceeds the corresponding threshold range, an in-depth analysis is conducted on the output of distributed energy and the charging and discharging strategy of the energy storage system to locate the anomaly and output the basis for parameter adjustment.

[0054] Based on the parameter adjustment criteria, the energy storage system's charging and discharging strategy is corrected and the distributed energy output allocation is optimized. The operation adjustment configuration is output, and the voltage stability index and power balance are recalculated. The support vector machine algorithm is used to predict the comprehensive performance of the two indicators and calculate the comprehensive index.

[0055] If the comprehensive index is greater than 0.9, the final optimized output is confirmed, which includes the corrected distributed energy output, energy storage charging and discharging strategy, comprehensive index, and the execution signal is distributed to resources in each stage.

[0056] The beneficial effects of this invention are:

[0057] 1. This invention discloses a source-grid-load-storage collaborative interactive optimization control method, which proposes a complete optimization control logic covering "scheduling-allocation-delay-protocol-closed loop" for complex business scenarios such as bidirectional regulation of power grid and distributed resources, communication delay and real-time feedback deviation, and cross-regional resource coordination.

[0058] 2. This invention first constructs a multi-objective optimization model based on spatiotemporal attributes (time dynamic weights + spatial clustering constraints), and integrates real-time operational data to generate a spatiotemporally adapted initial collaborative scheduling scheme. Then, through full-process interactive analysis and bidirectional adjustment capability assessment, it iteratively optimizes resource allocation ratios using linear programming and genetic algorithms, and dynamically corrects signal thresholds and edge response plans by combining delay-sensitive threshold calibration and compensation mechanisms (signal rearrangement, feedback cycle shortening). Based on this, it innovatively integrates the autonomous response mechanism with extended fields of the device communication protocol to generate an autonomous response plan supported by the protocol, and establishes a collaborative link between local response and wide-area optimization through wide-area resource collaborative parameter correction (cross-regional delay analysis, deviation threshold expansion). Finally, through aggregated evaluation of status indicators (voltage stability + power balance) and a closed loop of "signal distribution - continuous monitoring - dynamic adjustment," it achieves optimized system output.

[0059] 3. This invention effectively solves the problems of "rigid single target, lack of interaction in one-way commands, lagging edge response, protocol disconnect, and fragmented state assessment" in traditional solutions, and improves the collaborative interaction efficiency, power supply reliability and control accuracy of the source-grid-load-storage system. It can provide efficient and feasible technical support for energy management in complex power grid environments. Attached Figure Description

[0060] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0061] Figure 1 This is a flowchart illustrating the source-grid-load-storage collaborative interaction optimization control method in this embodiment. Detailed Implementation

[0062] The embodiments of the invention will now be described in detail with reference to the accompanying drawings.

[0063] This embodiment provides an optimized control method for the coordinated interaction of power generation, grid, load, and energy storage. It proposes a complete set of optimized control logic to address complex business scenarios such as bidirectional adjustment communication delays and real-time feedback deviations between the power grid and distributed energy sources, and cross-regional resource coordination. For example... Figure 1 As shown, the source-grid-load-storage coordinated interaction optimization control method includes the following steps:

[0064] Step S1: By collecting real-time operating data of distributed energy, power grid, load and energy storage equipment, and taking minimizing operating costs, maximizing power supply reliability and improving energy utilization efficiency as objective functions, a multi-objective optimization model is constructed. The model integrates relevant parameters of distributed energy output, flexible load regulation capability, energy storage system charging and discharging strategy and overall power grid operating status to obtain an initial collaborative scheduling scheme.

[0065] Step S1 in this embodiment is the foundation of the "source-grid-load-storage" collaborative interaction. Its core is to construct a multi-objective optimization model and generate an initial collaborative scheduling scheme. Addressing the shortcomings of two types of schemes in the background technology—"cloud-centralized systems lacking real-time response" and "edge-autonomous systems lacking wide-area collaboration"—this step solves the problems of traditional scheduling such as "limitations of a single objective, lack of inclusion of new energy uncertainties, and insufficient real-time performance," providing an initial decision-making benchmark for subsequent spatiotemporal expansion and bidirectional optimization steps. Specifically, the process of this step is as follows:

[0066] Step S11 involves acquiring real-time operational data from distributed energy sources, the power grid, loads, and energy storage devices via sensors to obtain the raw operational dataset. For example, the distributed energy source side collects power output data from photovoltaic arrays and wind turbine speed data; the power grid side monitors transformer load rate and line voltage fluctuations; the load side acquires peak demand data from industrial users and residential electricity consumption curves; and the energy storage devices record battery SOC status and charge / discharge efficiency. This real-time operational data is uploaded to the central server at a rate of seconds via the SCADA system (i.e., data acquisition and monitoring control system), forming a time-series dataset (structured data with timestamps) for subsequent optimization analysis.

[0067] Step S12: Extract relevant parameters from the original operational dataset, including distributed energy output, flexible load regulation capability, energy storage system charging and discharging strategy, and overall grid operation status, to obtain a parameter set. Distributed energy output refers to the power value predicted by an LSTM neural network based on meteorological data (e.g., light intensity, wind speed) and historical sequences, outputting a value every 15 minutes within the future scheduling cycle. This needs to consider the 10-20% fluctuation uncertainty caused by cloud cover (serving as the uncertainty boundary for the model input). Flexible load refers to interruptible or peak-shifting loads (e.g., industrial motors and air conditioning units). Regulation capability is calculated as base load minus adjustable capacity (e.g., a factory with a peak load of 500MW, of which 20% can be shifted to off-peak hours, resulting in a regulation capability of 100MW). The energy storage system's charging and discharging strategy is formulated based on SOC dynamic programming, setting upper and lower limits for SOC (e.g., 20%-80%), charging and discharging priorities (discharging is prioritized during peak hours to obtain higher electricity price revenue, and charging is done during off-peak hours to purchase electricity at lower prices), and superimposing time-of-use pricing (matching with the grid's peak and off-peak electricity price periods) and cycle life constraints (e.g., ≤2 charging and discharging times per day to avoid deep discharge accelerating aging). The overall grid operating status includes load factor (current load / total capacity), reserve capacity margin (adjustable power output - expected peak difference), and frequency deviation (monitored in real time via PMU, i.e., synchronous phasor measurement device); all parameters need to be normalized (e.g., mapped to the [0,1] interval) to eliminate dimensional differences and facilitate model calculation.

[0068] Step S13: Construct a multi-objective optimization model based on the relevant parameter set, integrating parameters related to distributed energy output, flexible load regulation capability, energy storage system charging and discharging strategy, and the overall grid operation status. The objective function for minimizing operating costs is defined as the total generation cost plus energy storage cycle losses, minus the peak-valley electricity price difference revenue. The objective function for maximizing power supply reliability is defined as minimizing the load cutoff probability, incorporating N-1 safety constraints (e.g., line overload threshold not exceeding 90%). The objective function for improving energy utilization efficiency is defined as maximizing the renewable energy penetration rate. The model uses NSGA-II (non-dominated sorting genetic algorithm) to solve the Pareto front (the set of non-dominated solutions for multi-objective optimization), setting the population size to 200 and the number of iterations to 500, and handling hard constraints (e.g., grid voltage deviation ≤ 5%) using a penalty function method.

[0069] Step S14: Solve the relevant parameter set through a multi-objective optimization model and obtain load demand forecast data from the relevant parameter set; if the load demand forecast data exceeds the preset threshold (e.g., peak load ≥ 800MW), then the flexible load adjustment capability (interruption / peak shift adjustable load) is used first to smooth the peak load; otherwise, the energy storage system charging and discharging strategy (e.g., peak discharge and valley charging) is used to adjust the supply and demand balance; finally, an initial collaborative scheduling scheme draft is generated (including photovoltaic power curtailment, energy storage charging and discharging amount, load peak shifting amount, standby unit activation instructions, etc.).

[0070] Step S15: Verify the consistency of the overall power grid operating status parameters for the initial draft of the coordinated dispatch scheme, and determine the final initial coordinated dispatch scheme.

[0071] This embodiment further expands upon the original "constructing a multi-objective optimization model → generating an initial collaborative scheduling scheme" by incorporating the dimensional expansion of time and space attributes. This addresses the deficiency of traditional initial schemes in "not considering spatiotemporal dynamics," improves the model's adaptability to complex power grid environments (such as peak-valley differences and regional load heterogeneity), and provides a more accurate initial decision-making benchmark for subsequent bidirectional optimization, delay compensation, and other steps. Specifically, step S1 in this embodiment also includes: step S16, expanding the time and space attributes of the variable dimensions to optimize the initial collaborative scheduling scheme.

[0072] More specifically, the process of step S16 is as follows:

[0073] Step S161: Extract relevant data of time attributes (such as the division of time periods in the next 24 hours, historical load data, and meteorological forecast data) and spatial attributes (such as the geographical partitioning of the power grid, the load density of each partition, and the proportion of distributed power sources) from the initial collaborative scheduling scheme, and organize them into a structured spatiotemporal dataset (including fields such as timestamp, partition ID, load density, and power source proportion).

[0074] Step S162 involves constructing a time-series-based dynamic weight allocation mechanism (i.e., dynamically adjusting the weights of corresponding terms in the objective function based on the load importance and fluctuation characteristics of different time periods) to improve the model's responsiveness to short-term fluctuations. For example, the next 24 hours are divided into 96 15-minute time segments. Based on historical load data and weather forecasts, a dynamic priority coefficient is assigned to each segment (for peak periods such as 8:00-11:00, the coefficient is set to 1.5 to prioritize power supply reliability; for off-peak periods such as 1:00-4:00, the coefficient is set to 0.8 to prioritize reducing operating costs). The time-series difference algorithm is used to calculate the load change rate for each time period, controlling the prediction error within ±3%. The weights of time-sensitive terms in the multi-objective optimization model (such as minimizing the operating cost objective function and maximizing the power supply reliability objective function) are adjusted based on the dynamic priority coefficient (e.g., the weight of maximizing the power supply reliability objective function during peak periods is increased from 0.3 to 0.45) to ensure that the coordinated scheduling scheme responds promptly to short-term fluctuations (such as sudden drops in photovoltaic output).

[0075] Step S163: For spatial attributes, such as dividing the power grid into 10 geographical zones, based on the load density of each zone (e.g., 2.5 MW / km² for zone A) and the proportion of distributed energy (e.g., 30% for zone A), the spatial load distribution characteristics are analyzed using the K-means clustering algorithm to generate a spatial correlation matrix (describing the load correlation between zones); based on the spatial correlation matrix, the upper limit of power flow transfer between geographical zones is calculated (e.g., the upper limit of power transfer between zone A and adjacent zones is 5 MW), and this is added as a spatial constraint to the multi-objective optimization model to limit the transmission loss caused by cross-zone scheduling (e.g., when the cross-zone power exceeds the upper limit, the model automatically reduces the output of that zone).

[0076] Step S164: The "dynamic weight allocation" of the time attribute and the "spatial correlation matrix + power transfer upper limit" of the spatial attribute are integrated into the multi-objective optimization model. A time smoothing term (e.g., penalizing drastic power fluctuations in adjacent time periods, with a weight of 0.2) and a spatial balance term (e.g., penalizing excessive load differences between sub-districts, with a weight of 0.15) are added to form a five-objective optimization model of "economy-reliability-efficiency-time smoothing-spatial balance". At the same time, hard constraints are added for cross-time period power fluctuation ≤10% (limiting the power change rate within 15 minutes) and cross-district voltage deviation <2% (limiting voltage differences between sub-districts).

[0077] Step S165 involves generating, for example, 1000 sets of spatiotemporal scenario data (covering combinations of different weather conditions and load fluctuations) through Monte Carlo simulation to evaluate the robustness of the expanded model. For instance, in the scenario "During peak hours, the load in zone A suddenly increases to 3.2MW / km² (from 2.5MW / km²)," the model automatically triggers: ① The neighboring zone (zone B) supports a 2MW power flow; ② Optimizes the scheduling strategy for the next four 15-minute segments (reducing the photovoltaic curtailment of zone A from 80MW to 70MW and increasing energy storage discharge from 200kWh to 250kWh), ensuring that the overall voltage deviation is <2% and the power balance is >0.95. Simultaneously, it integrates with weather warning services; if strong winds (affecting wind power output) are predicted for a zone within the next 6 hours, the model adjusts the reserve capacity ratio of that zone from 10% to 15% in advance, dynamically updating parameters through automated algorithms to generate a spatiotemporally adapted initial plan.

[0078] Step S166 involves a dual verification of the initial spatiotemporal adaptation scheme: verifying whether the "time smoothing term + spatial equalization term" is consistent with the five objective functions ("economy-reliability-efficiency-time smoothing-spatial equalization") (e.g., time smoothness ≥ 0.85, spatial equalization ≥ 0.9); and verifying whether the hard constraint of "cross-period power fluctuation ≤ 10% + cross-regional voltage deviation < 2%" is met. Upon successful verification, the final spatiotemporal adaptation initial collaborative scheduling scheme (including photovoltaic output, energy storage charging and discharging, and flexible load adjustment commands for each time period and region) is output.

[0079] Step S2 involves using an optimization algorithm to process the initial coordinated scheduling scheme or the final spatiotemporally adapted initial coordinated scheduling scheme, performing iterative calculations for the entire energy interaction process, and determining the final resource allocation ratio under the bidirectional optimization logic, where the bidirectional optimization logic refers to the mutual adjustment mechanism between the power grid and distributed energy.

[0080] Step S2 in this embodiment is the core optimization step of the "source-grid-load-storage" coordinated interaction. It follows the initial coordinated scheduling scheme generated in step S1 or the final initial coordinated scheduling scheme with spatiotemporal adaptation. Through iterative calculation across all stages and bidirectional optimization logic (the mutual adjustment mechanism between the power grid and distributed energy resources), it overcomes the shortcomings of traditional scheduling, which involves "one-way commands and lack of interaction." Ultimately, it outputs a resource allocation ratio that balances resource efficiency, system stability, and economy, providing a precise resource allocation benchmark for subsequent edge control, delay compensation, and other steps. Specifically, the process of this step is as follows:

[0081] Step S21: Obtain the initial coordinated scheduling scheme or the final spatiotemporally adapted initial coordinated scheduling scheme (including instructions for photovoltaic output, energy storage charging and discharging, flexible load adjustment, etc. for each time period and region), and parse and extract the energy interaction data and initial resource allocation ratio of each link. The energy interaction data for each link includes grid load demand (e.g., a total grid load demand of 1000MW in the initial scheme) and total distributed energy supply capacity (e.g., a total supply of 400MW from photovoltaic and wind power). The initial resource allocation ratio includes the load proportion borne by the grid (e.g., 70%) and the load proportion borne by distributed energy (e.g., 30%). Organize the above data into a structured basic dataset (including fields: link ID, energy type, initial allocation ratio, supply capacity, etc.) for subsequent analysis.

[0082] Step S22: Based on the basic dataset, perform dynamic analysis of energy interaction across all stages (covering the entire process of "source-grid-load-storage"). Through hierarchical calculation (breaking down the process into "power source → grid → load → energy storage"), the interaction intensity between stages is calculated (quantifying the degree of mutual influence between each stage, such as the sensitivity of grid load changes to distributed photovoltaic output, calculated as "photovoltaic output change rate / grid load change rate") and adjustment demand (deriving the resource adjustment demand of each stage based on the interaction intensity, such as when the grid load needs to be reduced from 1000MW to 900MW, distributed energy needs to increase its supply by 100MW).

[0083] Step S23, based on the interaction intensity and regulation demand, combined with the two-way optimization logic (the mutual regulation mechanism between the power grid and distributed energy), evaluates the power grid regulation capacity (evaluates the power grid's ability to adjust the load through methods such as "activation of standby units and load switching", for example, the maximum adjustable load of the power grid is 200MW) and the distributed energy response capacity (evaluates the ability of distributed energy such as photovoltaic and wind power to respond to regulation demand through methods such as "output adjustment and charge-discharge switching", for example, the maximum increase in supply of distributed energy is 150MW).

[0084] Step S24 involves iteratively processing the initial resource allocation ratio, optimizing the balance between the grid's regulation capacity and the distributed energy response capacity through algorithms. For example, a linear programming algorithm can be used with the objective function of "minimizing the deviation between the grid's peak load and the distributed energy supply," under the constraints of "grid load ≥ 800MW, distributed energy supply ratio ≤ 40% of total demand." This yields the initial resource allocation ratio (e.g., the grid handles 700MW, distributed energy handles 300MW, with a deviation of 100MW). Then, for the entire energy interaction process, a genetic algorithm (population size 50, iterations 100) is used for iteration. The initial resource allocation ratio is adjusted through "crossover and mutation." The evaluation function is a weighted sum of "energy utilization rate (weight 0.6) + system stability (weight 0.4)," ultimately outputting an optimized allocation scheme (e.g., 720MW for the grid, 280MW for distributed energy, energy utilization rate 85%, system stability index 0.9).

[0085] Step S25: For the optimized allocation scheme, verify whether it meets the energy interaction balance requirements of the entire process (e.g., the deviation between grid load and distributed energy supply ≤ 50MW, and power fluctuation of each process ≤ 10%). If the balance requirements are met, output the final resource allocation ratio and proceed directly to step S26; if the balance requirements are not met, recalculate the resource allocation ratio based on the bidirectional optimization logic using a dynamic game model (Nash equilibrium). For example, with the objective of "lowest total regulation cost", input the grid regulation cost (e.g., 0.5 yuan / MW) and the distributed energy regulation cost (e.g., 0.3 yuan / MW) to obtain the final resource allocation ratio (e.g., the grid bears 73%, i.e., 730MW, and distributed energy bears 27%, i.e., 270MW, with a total regulation cost of 110 yuan).

[0086] Step S26: Based on the final resource allocation ratio, generate resource configuration instructions for each stage (e.g., "activate standby units of 30MW in the power grid", "increase output of distributed photovoltaic power by 70MW", "discharge energy storage of 50kWh"); at the same time, record the execution data of coordinated scheduling (e.g., instruction issuance time, resource adjustment amount), and confirm the stable operation of the entire system through system stability assessment (e.g., voltage deviation ≤5%, frequency deviation ≤0.2Hz).

[0087] Step S27: Based on the final resource allocation ratio (e.g., the power grid accounts for 73% and distributed resources account for 27%), combined with the topology and node capability profiles of the edge computing network (e.g., the computing resources, bandwidth, and load rate of nodes A / B / C), a load balancing algorithm maps the macro-level resource allocation ratio to specific execution instructions for each edge node. The mapping logic involves allocating the total share of distributed resources (27%) to each edge node proportionally based on the historical load factor (e.g., 1.2) and current available resources. The final output is an edge node resource allocation dataset (e.g., node A accounts for 30% of the total resources, node B for 25%, and node C for 45%), which serves as input for the subsequent step S3.

[0088] Step S3: Extract key variables from the final resource allocation ratio, generate control signals, and obtain an edge distribution response plan containing the control release signal sequence through real-time feedback from local devices and global load balancing optimization of edge nodes.

[0089] Step S3 in this embodiment is the edge deployment stage of the "source-grid-load-storage" collaborative interaction. Following the final resource allocation ratio output in step S2, it solves the shortcomings of traditional scheduling—"lagging edge device response and lack of autonomous adjustment"—through a closed-loop process of "key variable extraction → control signal generation → local real-time feedback → edge response plan generation." Ultimately, it outputs an edge-distributed response plan adapted to edge computing, providing an edge-side execution benchmark for subsequent steps such as latency compensation and protocol fusion. Specifically, the process of this step is as follows:

[0090] Step S31: Extract core variables (resource percentage of each node) from the final resource allocation ratio (e.g., node A accounts for 30%, node B accounts for 25%, and node C accounts for 45% in the edge computing network), and calculate the resource occupancy weight using a weighted average method (incorporating historical load characteristics). The calculation formula is: Weight = Current percentage × Historical load factor (e.g., assuming the historical load factor is 1.2, node A weight = 30% × 1.2 = 36%, node B = 25% × 1.2 = 30%, and node C = 45% × 1.2 = 54%). Store the weights in the local database to form a preliminary variable set.

[0091] Step S32: Based on the initial variable set, a control signal is generated using a time series analysis algorithm. For example, in an edge computing network where node A accounts for 30%, node B for 25%, and node C for 45%, the weights are mapped to a signal strength range of 0-100, using the formula: Signal Strength = Weight × 100 / Maximum Weight. Then, the values ​​of node A ≈ 66.67, node B ≈ 55.56, and node C = 100 are calculated, forming a control signal release sequence [66.67, 55.56, 100]. The signals are sorted from highest to lowest strength (node ​​C → node A → node B) and released to each edge node in timestamp order via a message queue.

[0092] Step S33: The control signal is transmitted to the local device (e.g., the edge network node controller) to perform status judgment and frequency adjustment. Status judgment involves acquiring local device operating status data such as current resource utilization and load rate to determine if it falls within a preset response range (e.g., resource utilization deviation ≤ 5%). Frequency adjustment involves adjusting the signal sequence transmission frequency (e.g., increasing from 1 time / second to 2 times / second) if the operating status data exceeds a preset threshold (e.g., node A's load rate exceeds 80%). Based on these adjustments, the adjusted control signal is generated.

[0093] Step S34 involves using the processed control signal to drive the local device to perform real-time feedback, including strategy adjustment and feedback verification, and returning real-time feedback data. Strategy adjustment involves the device automatically adjusting resource allocation based on a preset threshold, such as signal strength ≥ 70, after receiving the signal. For example, if the signal strength of node C is 100 > 70, the resource allocation increases to 50%; if the signal strength of node A is 66.67 < 70, the resource allocation decreases to 28%, executed via a local embedded script. Feedback verification involves collecting device feedback data, such as the adjusted resource ratio, and determining whether it meets the expected response standard, such as a resource ratio deviation ≤ 3%.

[0094] Step S35: Based on real-time feedback data, analyze the correspondence between signal strength and resource adjustment amount, generate autonomous response execution rules (e.g., if the signal strength is ≥70, increase resource allocation by 5%; otherwise, decrease it by 2%), and output the response mapping path.

[0095] Step S36: Based on the response mapping path, evaluate and verify the load balancing of the edge distribution scheme. The load balancing is calculated using the formula: Load Balancing = 1 - (Standard deviation of resource share for each node / Average share). For example, after adjustment, nodes A = 28%, B = 23%, and C = 49%, with a standard deviation of 0.13 and an average share of 0.33, resulting in a balance of approximately 0.61. During load balancing verification, if the load balancing is ≥ 0.5 (preset threshold), the edge distribution scheme meets the standard, and an optimized distributed response framework is generated; otherwise, the edge distribution scheme is re-optimized until the load balancing is ≥ 0.5 (preset threshold), generating an optimized distributed response framework.

[0096] Step S37: Based on the optimized distributed response framework, generate the final edge distributed response plan, which includes a sequence of control and release signals. First, integrate the updated data of resource allocation (e.g., node C has 50% of resources, node A has 28%), determine the execution cycle (e.g., iterate once every 5 minutes), and then integrate the adjustment results of each node through a distributed consensus algorithm (e.g., Raft). After verifying that the global balance is ≥0.5, it is sent to each edge node for execution through the API interface, forming a closed-loop control logic.

[0097] Step S4: Evaluate the communication delay level and the deviation of the instantaneous feedback response of the control and release signal sequence, integrate the delay compensation mechanism, and determine the adjustment threshold of the delay-sensitive signal.

[0098] Step S4 in this embodiment is the latency perception and threshold calibration stage of the "source-grid-load-storage" collaborative interaction. It follows the edge distribution response plan (including control signals) output in step S3, and through a closed-loop process of "latency monitoring → compensation adjustment → threshold calibration → process stabilization," it addresses the deficiency in traditional scheduling where "communication latency and real-time feedback deviation are not co-optimized." Ultimately, it outputs a latency-sensitive signal to adjust the threshold, providing a latency perception decision-making benchmark for subsequent steps such as wide-area parameter correction and protocol fusion. Specifically, the process of this step is as follows:

[0099] Step S41: For the control and release signal sequence (e.g., signal strength 100 at edge node C, ≈66.67 at node A), collect signal transmission data through the real-time monitoring module (e.g., release 100 signals within 1 second, recording the "send → receive" delay time for each signal), and statistically analyze the delay data and delay level classification. When analyzing the delay data, calculate the average communication delay (e.g., 0.05 seconds) and standard deviation (e.g., 0.01 seconds). When classifying the delay level, based on statistical analysis (e.g., a normal distribution model), determine whether the delay level meets the system requirements (preset threshold, such as 0.06 seconds), and classify the delay level status into two levels: "normal" (≤0.06 seconds) and "high delay" (>0.06 seconds).

[0100] Step S42: Based on the latency level classification, invoke the preset latency compensation mechanism. Under this mechanism, for the signal transmission sequence in the "high latency state", the system executes a priority rearrangement strategy - that is, sorts them from high to low according to the latency duration (the longer the latency, the higher the priority), and outputs the adjusted signal transmission sequence (for example, raising the priority of the signal with a latency of 0.07 seconds to the highest, giving priority to retransmission or channel switching).

[0101] Step S43: Based on the adjusted signal release sequence, collect real-time feedback response data (e.g., the average deviation of the feedback time from the expected time is 0.02 seconds), and perform deviation threshold judgment and feedback cycle optimization. Specifically, during deviation threshold judgment, if the feedback deviation exceeds a preset range (e.g., >0.03 seconds), the feedback cycle is dynamically adjusted (e.g., shortened from 1 second / cycle to 0.5 seconds / cycle) through feedback frequency adjustment parameters (e.g., a PID controller). During feedback cycle optimization, the optimized feedback cycle (e.g., 0.5 seconds) is output to ensure the timeliness of the feedback data.

[0102] Step S44: Extract key time nodes (such as peak delay time, maximum deviation time) from the optimized feedback cycle, and combine them with delay-sensitive characteristics (such as delay ≤0.05 seconds required in industrial control scenarios) to determine whether the delay-sensitive signal adjustment threshold needs to be dynamically updated. If the delay of the key time node is close to the preset threshold (such as 0.055 seconds ≈ 0.06 seconds), adjust the signal threshold range (such as widening it from 0.06 seconds to 0.065 seconds, or tightening it to 0.055 seconds), and output the initial signal adjustment threshold range (such as 0.055~0.065 seconds).

[0103] Step S45: Obtain the initial signal adjustment threshold range. Combine this with historical data on communication latency and response deviation (e.g., 5% of the past 1000 signals had latency exceeding the standard), and use a Support Vector Machine (SVM) algorithm to calibrate the threshold range. For example, with the goal of "minimizing the latency exceeding the standard rate," the SVM is trained on historical data to output the final latency-sensitive signal adjustment threshold (e.g., the optimized threshold is adjusted from 0.06 seconds to 0.055 seconds to adapt to high-load scenarios).

[0104] Step S46: Based on the delay-sensitive signal adjustment threshold, the control and release signal sequence is optimized in real time. If the delay of the control and release signal sequence exceeds the delay-sensitive signal adjustment threshold (e.g., 0.056 seconds > 0.055 seconds), the delay compensation mechanism is activated again, for example, reducing the signal transmission frequency from 100 times / second to 80 times / second to reduce network congestion. The delay level is continuously monitored until the delay stabilizes at or below the final delay-sensitive signal adjustment threshold, and a stable signal release process (including transmission frequency, compensation rules, and threshold triggering conditions) is output.

[0105] Step S5: Integrate the delay-sensitive signal adjustment threshold into the correction process of the wide-area resource collaboration parameters to obtain the wide-area resource interaction correction parameters and signal release adjustment scheme.

[0106] Step S5 in this embodiment is the wide-area parameter correction stage for the collaborative interaction of "source-grid-load-storage". Following the delay-sensitive signal threshold adjustment output in step S4, it solves the problem of "disconnection between delay threshold and wide-area collaborative parameters" in traditional scheduling through a closed-loop process of "threshold judgment → cross-regional delay analysis → response deviation expansion → wide-area parameter correction". Finally, it outputs expanded interactive correction parameters, providing a decision-making benchmark for wide-area collaboration in subsequent steps such as protocol fusion and autonomous response plan generation. Specifically, the process of this step is as follows:

[0107] Step S51: Extract delay-sensitive feature data (such as threshold value and delay fluctuation range) from the delay-sensitive signal adjustment threshold (e.g., the current threshold of 15ms), and analyze to obtain the initial threshold data for signal adjustment (i.e., 15ms).

[0108] Step S52: Compare the initial threshold data (15ms) with a preset standard (e.g., 10ms). If the initial threshold > the preset standard (15ms > 10ms), cross-regional latency distribution analysis is automatically triggered to obtain communication latency data (e.g., cross-regional latency distribution mean 12ms, standard deviation 3ms), and output the cross-regional latency distribution range (e.g., 9~15ms, calculated based on mean ± standard deviation). For example, consider the following three geographical zones: Zone A (urban microgrid): load density 2.5MW / km², distributed photovoltaic (PV) 30% (local power source is mainly PV); Zone B (suburban wind farm): load density 0.5MW / km², distributed wind power 80% (local power source is mainly wind power); Zone C (industrial park): load density 3.0MW / km², no distributed power source (completely dependent on external power). The communication delay data for zones A→B, B→C, and A→C are collected (transmitted via SCADA system or 5G private network). The following calculations are made: Average cross-regional delay = 12ms (e.g., wind power command delay from A→B, load response delay from B→C average 12ms); Standard deviation of cross-regional delay = 3ms (delay fluctuates across different links, e.g., photovoltaic command delay from A→C sometimes reaches 15ms, sometimes only 9ms). The cross-regional delay distribution range is output: based on "mean ± standard deviation", it is 9~15ms (i.e., delay fluctuates between 9ms and 15ms). If the initial threshold ≤ preset standard (e.g., 10ms = 10ms), automatic cross-regional delay distribution analysis is not triggered. Instead, the response deviation value calculated based on the cross-regional delay distribution (e.g., 2ms calculated based on the average delay of 12ms) is used to proceed to step S55.

[0109] Step S53: Calculate the response deviation value of the instantaneous feedback based on the cross-regional delay distribution interval (e.g., based on an average delay of 12ms, the response deviation is calculated to be 2ms). If the response deviation value is ≥ the upper limit of the response deviation evaluation threshold (e.g., the upper limit of the evaluation threshold is 2ms), proceed to step S54; otherwise, jump directly to step S55.

[0110] Step S54: For response deviation values ​​that reach the upper limit of the evaluation (e.g., a response deviation of 2ms calculated based on the cross-regional delay average of 12ms, and ≥ the upper limit of the evaluation threshold of 2ms), the response deviation evaluation threshold is calculated by expanding the formula based on the standard deviation of the cross-regional delay distribution (e.g., 3ms)—that is: New response deviation evaluation threshold = Original response deviation evaluation threshold + Cross-regional delay standard deviation × 1.5. Taking the original evaluation threshold of 10ms and the cross-regional delay standard deviation of 3ms as an example, the calculated new response deviation evaluation threshold = 10 + 3 × 1.5 = 14.5ms, which is output as the expanded response deviation evaluation data.

[0111] Step S55: Based on the response deviation value or response deviation assessment data (e.g., 14.5ms), and combined with the wide-area resource allocation situation (e.g., cross-regional energy allocation ratio, inter-node delay fluctuation range), adjust the weight value of the wide-area resource coordination parameter (e.g., the original coordination parameter weight is 0.8, and the weight is adjusted according to the ratio of the new response deviation assessment threshold to the preset standard 14.5 / 10=1.45); output the adjusted coordination parameter (e.g., 0.8×1.45=1.16).

[0112] Step S56: Based on the adjusted collaboration parameters (e.g., 1.16) and combined with business scenario requirements (e.g., stability requirements for cross-regional data synchronization), generate specific values ​​for the interaction correction parameters (e.g., optimized setting to 1.2); output the final wide-area resource interaction correction parameters (i.e., 1.2), and verify them through simulation tests (e.g., when the inter-node latency fluctuation is ≤5ms, the packet loss rate drops from 0.8% to below 0.5%) to ensure the stability of the parameters in multi-node collaboration.

[0113] Step S57: Using wide-area resource interaction correction parameters (e.g., 1.2), optimize the threshold adjustment process (e.g., threshold judgment, compensation mechanism) for delay-sensitive signals. Adjust the delay tolerance range for signal processing (e.g., extend it from 10ms to 14.5ms); optimize the cross-regional signal synchronization strategy (e.g., add backup channel switching logic); finally output the optimized signal broadcasting adjustment scheme, forming a complete logical chain of "threshold detection → parameter correction → performance verification".

[0114] Step S6: Based on the wide-area resource interaction correction parameters and signal release adjustment scheme, integrate the autonomous adjustment parameter update path of the autonomous response mechanism with the message triggering and response interface of the device communication protocol, and determine the response coordination scheme under the embedded protocol extension field configuration.

[0115] Step S6 in this embodiment is the protocol fusion and coordination stage for the collaborative interaction of "source-network-load-storage". Following the wide-area resource interaction correction parameters and signal release adjustment scheme output in step S5, it solves the defect of "disconnection between autonomous response mechanism and device communication protocol" in traditional scheduling through a closed-loop process of "communication interface analysis → protocol field mapping → autonomous adjustment path construction → interactive configuration synchronization". Finally, it outputs a response coordination scheme embedding protocol extension fields, providing a protocol-level collaborative benchmark for subsequent steps such as autonomous response plan generation and cloud deviation correction. Specifically, the distance flow of this step is as follows:

[0116] Step S61: Obtain the communication interface data stream (including message trigger time, data packet content, and protocol type) of the edge device (e.g., smart meter, energy storage controller). Classify the message trigger time and content in the communication interface data stream. Classify by "trigger type": periodic trigger, event trigger, and abnormal trigger; classify by "content priority": control command > status feedback > heartbeat packet. Output preliminary message trigger classification results (e.g., "event trigger - control command" has the highest priority, and "periodic trigger - heartbeat packet" has the lowest priority).

[0117] Step S62: Based on the message trigger classification results, parse the extended field structure of the device communication protocol (e.g., the ASN.1 extended field of the IEC61850 protocol, the function code extended field of Modbus RTU), and perform field-trigger mapping and priority sequence judgment. Specifically, during field-trigger mapping, determine the corresponding mapping relationship between extended fields and message triggers (e.g., "Event Trigger - Control Command" corresponds to the extended field "Priority Identifier," and "Periodic Trigger - Status Feedback" corresponds to the "Timestamp Field"). During priority sequence judgment, determine the priority sequence of the field configuration based on the "service urgency" of the mapping relationship (e.g., Control Command > Status Feedback) (e.g., "Priority Identifier" > "Timestamp Field" > "Reserved Field").

[0118] Step S63: Based on the priority sequence configured in the fields, integrate the adjustment logic of the autonomous response mechanism (e.g., "increase response priority when load exceeds 80%)" and the signal release adjustment scheme to construct a dynamic path for parameter updates. For example, a dynamic weighted algorithm is used to adjust the response priority weight. The initial weight is 0.5, and it is updated according to the formula: New weight = Initial weight × (1 + Load rate / 100). For example, when the load rate is 90%, the new weight = 0.5 × 1.9 = 0.95. At the same time, the "delay tolerance range, such as 14.5ms" in step S5 is synchronized to the dynamic path (as a constraint condition for weight adjustment). Output the initial path adjustment scheme (including the mapping rules of "load rate → weight → priority" + delay tolerance constraints).

[0119] Step S64: Using the initial path adjustment scheme, the interaction configuration of the device communication interface (such as polling period, data packet size threshold, retransmission mechanism) is adjusted in real time. For example, the polling period of "event trigger - control command" is shortened from 10ms to 5ms, and "periodic trigger - heartbeat packet" is extended from 1s to 2s. The execution order of the response mechanism is determined according to the priority sequence of the field configuration (such as the "priority flag" field is processed first, and the "time stamp field" is processed second).

[0120] Step S65: Collect actual interaction data of the device communication interface (such as message trigger time and field filling content), compare it with the interaction configuration, and identify inconsistencies (such as "Event Trigger - Control Command" not polling at 5ms); for the inconsistencies, analyze the cause (such as network congestion causing polling delay) through data comparison algorithms (such as CRC check and field matching); output the adjusted execution strategy (such as widening the polling period of "Event Trigger - Control Command" to 8ms, or switching to the backup channel).

[0121] Step S66: According to the adjusted execution strategy, update the interaction configuration of the device communication interface (such as polling period and retransmission threshold), and synchronously write the updated configuration into the protocol extension field (such as setting the value of "priority flag" to 0.95 and "polling period field" to 8ms); verify the consistency between the extension field and the interaction configuration (such as the field value matching the actual polling period); output the response coordination scheme (including field mapping rules, interaction configuration parameters, and execution order) under the embedded protocol extension field configuration.

[0122] Step S7: Use a response coordination scheme to process the local resource allocation execution unit in the edge distribution response plan, generate an adjustment sequence of the integrated device communication protocol delay management rule set, and obtain the autonomous response plan supported by the protocol.

[0123] Step S7 in this embodiment is the edge resource deployment and protocol solidification stage of the "source-network-load-storage" collaborative interaction. Following the response coordination scheme with embedded protocol extension fields output in step S6, it solves the defect of "disconnect between edge resource allocation and protocol delay management" in traditional scheduling through a closed-loop process of "edge node status awareness → dynamic resource allocation → protocol delay rule integration → autonomous response plan generation." Ultimately, it outputs an autonomous response plan supported by the protocol, providing an executable baseline for subsequent cloud-based deviation correction and continuous monitoring. Specifically, the process of this step is as follows:

[0124] Step S71: Obtain real-time status data of distributed nodes (such as remaining computing resources, bandwidth utilization, and current load rate) from the edge distribution environment (such as an edge computing network with 10 nodes), and determine the priority ranking of key nodes through communication load analysis (such as data transmission volume between nodes and message queue length). (For example, nodes with a load rate > 0.7 are "high priority" and should be given priority in resource allocation.)

[0125] Step S72: Based on the priority of key nodes, obtain the availability information of local resources (e.g., initial computing resources of 1000 units and bandwidth of 500Mbps for each node); if the resource utilization rate of a certain node is greater than the preset threshold (e.g., load rate > 0.8), then trigger dynamic adjustment: transfer the excess tasks to nodes with a load rate < 0.5, prioritize high-priority tasks during the transfer process (e.g., high-priority tasks account for 40%), and ensure that the task migration delay is ≤ 50ms; output a preliminary resource allocation plan (e.g., after adjustment, the load rate of each node is stable between 0.6 and 0.75, and the overall resource utilization rate is improved by about 15%).

[0126] Step S73: For the preliminary resource allocation plan, collect the interaction logs of the edge devices (such as protocol version number, data packet format, and handshake failure records), analyze the communication protocol compatibility between devices by comparing protocol versions (such as the field differences between ModbusRTUv1.0 and v2.0), and determine the direction of protocol adaptation and optimization (such as uniformly upgrading to ModbusRTUv2.0 to support extended fields).

[0127] Step S74: Based on the protocol adaptation and optimization direction, obtain the latency management history (such as the latency peak period and packet loss rate data of the past 30 days); if the latency fluctuation is greater than the predetermined range (such as the latency peak is greater than 30ms and mostly occurs during the traffic peak period of 18:00-20:00), then adjust the transmission parameters of the communication protocol (such as automatically switching to the backup channel when the latency exceeds 30ms, and reducing the data packet size from 1024 bytes to 512 bytes); pre-allocate 20% of additional bandwidth resources for the peak period (18:00-20:00); output the latency optimization rule set (including latency trigger threshold, channel switching logic, and bandwidth pre-allocation strategy).

[0128] Step S75: Drive the edge device to execute using a delay optimization rule set and collect execution feedback of rule adjustment (such as the delay value after switching to the backup channel and the packet loss rate after the data packet is reduced). By analyzing the abnormal data in the feedback (such as the failure of backup channel switching), determine the triggering conditions for autonomous response (such as triggering backup channel switching when "delay > 30ms and main channel packet loss rate > 5%)", and output the response mechanism configuration scheme (including triggering conditions, execution actions, and fallback logic).

[0129] Step S76: Based on the response mechanism configuration scheme, obtain the compatibility data supported by the protocol (such as the matching degree between extended fields and device firmware, and the adaptability of protocol version and rules), and verify it in real time through device interaction (such as sending test commands to nodes to check the response time and the correctness of protocol field parsing). After confirming that there are no errors, generate the autonomous response plan supported by the protocol. The autonomous response plan includes: ① a resource allocation strategy based on load balancing (such as high / medium / low priority tasks accounting for 40% / 35% / 25%); ② an adjustment sequence integrating the latency management rule set (such as switching channels when latency > 30ms, pre-allocating 20% ​​bandwidth during peak periods); ③ a machine learning-driven traffic prediction module (such as inputting traffic data from 288 sampling points in the past 24 hours and node load rate to predict traffic in the next hour with an error ≤ 5%). Verification indicators for the autonomous response plan: such as response time < 40ms, overall response efficiency improvement of 18%, and system stability improvement of 10% under burst traffic.

[0130] Step S8: Combine the wide-area resource interaction correction parameters and generate the final interaction scheme based on the autonomous response plan.

[0131] Step S8 in this embodiment is the wide-area deviation correction stage for the collaborative interaction of "source-grid-load-storage". Following the protocol-supported autonomous response plan output in step S7, it addresses the deficiency of "disconnect between edge autonomous planning and cloud optimization paradigm" in traditional scheduling through a closed-loop process of "deviation detection → parameter correction → model verification → scheme iteration". Finally, it outputs the corrected interaction scheme, providing a wide-area collaborative optimization benchmark for subsequent steps such as status indicator evaluation and continuous monitoring. Specifically, the process of this step is as follows:

[0132] Step S81: Collect real-time data from the autonomous response plan (e.g., the cross-regional energy allocation ratio on the edge side) and the cloud optimization paradigm (e.g., the regional energy allocation benchmark preset in the cloud) through the data analysis system, and calculate the deviation between the two (e.g., a preset threshold of 5.0% and an actual detected deviation of 7.2%). If the deviation value > the preset threshold (e.g., 7.2% > 5.0%), the deviation data is marked as an object to be processed, and historical data (e.g., similar deviation records from the past 30 days) and real-time monitoring data (e.g., the current cross-regional power flow) are integrated to form an initial dataset with deviation. If the deviation value ≤ the preset threshold (e.g., 4.5% ≤ 5.0%), the original wide-area resource interaction correction parameters (e.g., 1.2) are directly retained, and real-time monitoring data (e.g., the current cross-regional power flow) is integrated to form an initial dataset without deviation.

[0133] Step S82: For either a biased initial dataset (containing the biased objects to be processed) or an unbiased initial dataset, analyze the sources of bias or verify the parameter benchmarks, and use an optimization algorithm (such as a genetic algorithm) to analyze the wide-area resource interaction correction parameters (including wide-area resource coordination parameters, such as cross-regional energy allocation indicators); set the algorithm parameters, with an initial population size of 100 and 50 iterations; define the fitness function, with the objective of "minimizing energy allocation bias (e.g., weight 0.6) + maximizing resource utilization (e.g., weight 0.4)"; output the adjusted set of wide-area resource coordination parameters (e.g., the cross-regional energy allocation indicator is corrected from 25% to 28.5%).

[0134] Step S83: Based on the wide-area resource coordination parameter set (e.g., a cross-regional energy allocation indicator of 28.5%), and combined with the cross-regional energy allocation indicator in the wide-area resource interaction correction parameters (e.g., region A should allocate 30% and region B should allocate 25%), the resource coordination parameters are further refined and decomposed. For example, the indicator is split according to "regional load density + distributed power generation ratio". Region A has a high load density, so the allocation ratio is increased from 25% to 28.5%, while region B remains at 25%. The regional energy allocation scheme is output (including power output, load regulation, and energy storage charging and discharging instructions for each region).

[0135] Step S84: Based on the regional energy allocation scheme, verify whether the resource allocation meets the preset balance conditions of wide-area resource coordination (e.g., cross-regional power fluctuation ≤8%, line overload rate ≤90%). If the balance conditions are met, proceed directly to step S85. If the balance conditions are not met (e.g., power fluctuation in region A reaches 10%), fine-tune the allocation indicators (e.g., reduce the allocation ratio of region A from 28.5% to 27%), and output the balanced allocation result.

[0136] Step S85: Combine the balanced allocation results with the wide-area resource interaction correction parameters (e.g., latency sensitivity threshold of 1.2) to generate a preliminary interaction scheme (including scheduling instructions and latency compensation rules for each region); verify whether the scheme meets the constraints of the cloud optimization paradigm (e.g., voltage deviation ≤ 5%, frequency deviation ≤ 0.2Hz, energy allocation deviation ≤ 5%); if it does not meet the constraints (e.g., energy allocation deviation still reaches 6%), record the non-compliant constraints (e.g., "energy allocation deviation in region A exceeds the standard"), and form the data of the scheme to be optimized.

[0137] Step S86: For the data of the scheme to be optimized (including non-compliant constraints), use an optimization algorithm (such as particle swarm optimization) to locally adjust the constraints; with the goal of "minimizing the constraint deviation", iteratively optimize the energy allocation ratio of region A (adjusted from 27% to 28.2%); repeatedly verify the constraints until all constraints meet the standards, and output the final interactive scheme (e.g., region A allocates 28.2%, region B allocates 25%, and the energy allocation deviation is ≤5%).

[0138] Step S9: Obtain the status index data in the final interactive scheme, aggregate and evaluate the overall effect of distributed energy output and energy storage system charging and discharging, obtain the final optimized output, and execute signal distribution to resources in each link to realize continuous monitoring and response loop under the expanded control range of dynamic adjustment from local to wide area.

[0139] Step S9 in this embodiment is the state aggregation assessment and final optimization stage of the "source-grid-load-storage" collaborative interaction. Following the final interaction scheme output in step S8, it solves the shortcomings of traditional scheduling, such as "fragmented state assessment and lack of closed-loop optimization" and "fixed control range and lack of continuous feedback," through a closed-loop process of "state index extraction → threshold comparison → anomaly analysis → parameter correction → comprehensive prediction → continuous monitoring and response." The final output is an optimized result that balances voltage stability and power balance, ultimately achieving "continuous monitoring and response loop under the expanded control range of dynamic adjustment from local to wide-area." Specifically, the process of this step is as follows:

[0140] Step S91: Extract status indicator data (including "voltage stability" and "power balance") from the final interactive solution and classify them according to indicator type. Voltage stability data includes the actual voltage value of each node (e.g., the voltage of a node is 1.02 times the rated value) and the rated voltage value. Power balance data includes distributed energy output (e.g., 200kW photovoltaic, 150kW wind power), energy storage system charging and discharging power (e.g., 50kW), and total load value (e.g., 400kW). Output the set of classified indicators and store it in the cloud database (for subsequent analysis and traceability).

[0141] Step S92: Based on the set of classification indicators, compare the voltage stability assessment and power balance assessment using preset threshold ranges (e.g., voltage stability in power systems typically requires a deviation of ≤5%, and power balance typically requires a deviation of ≤10%). For voltage stability assessment, calculate the voltage stability index using the formula: 1 - |Actual Voltage - Rated Voltage| / Rated Voltage. For example, if the index at a certain node is 0.98 (>0.95, meeting the stability standard), the index is calculated. For power balance assessment, calculate the power balance using the formula: (e.g., Balance = 1 - |Total Output - Total Load| / Total Load). In the example, the balance is 0.875 (<0.9, indicating room for optimization). Output the preliminary assessment results, such as marking "Voltage Stable" or "Power Balance Needs Optimization".

[0142] Step S93: If the preliminary assessment results show that a certain indicator exceeds the corresponding threshold range (e.g., power balance 0.875 < 0.9), then conduct in-depth analysis on the distributed energy output data (photovoltaic, wind power) and the energy storage system's charging and discharging performance (charging and discharging power, SOC): identify abnormal fluctuation ranges (e.g., photovoltaic output fluctuates by 15% between 10:00 and 10:15); combine the operating logs of the distributed energy and energy storage systems (e.g., photovoltaic inverter temperature records, energy storage battery charge and discharge cycles) to obtain the corresponding changes in operating parameters (e.g., inverter temperature rises from 40℃ to 55℃, energy storage SOC drops from 70% to 55%); determine whether there is uneven equipment load (e.g., 3 inverters in the photovoltaic array have a load rate > 90%, while the rest are < 50%), and output the basis for parameter adjustment (e.g., "uneven inverter load leads to output fluctuations", "low energy storage SOC leads to unstable discharge power").

[0143] Step S94: Based on the parameter adjustment criteria, perform energy storage system charging and discharging strategy correction and distributed energy output allocation optimization. Specifically, during energy storage system charging and discharging strategy correction, dynamic planning based on SOC is used to adjust the charging and discharging power (e.g., reducing the energy storage discharging power from 50kW to 40kW to avoid excessively low SOC). During distributed energy output allocation optimization, the photovoltaic inverter load is balanced (e.g., reducing the output of high-load inverters from 80kW to 70kW, and increasing the output of low-load inverters from 50kW to 60kW). The output operation adjustment configuration includes the corrected energy storage charging and discharging power and distributed energy output allocation ratio.

[0144] Step S95: Based on the operational adjustment configuration, recalculate the voltage stability index (e.g., increasing it from 0.98 to 0.99) and power balance (e.g., increasing it from 0.875 to 0.9). Use a Support Vector Machine (SVM) algorithm to predict the combined performance of these two indicators: calculate the combined index using a weighted average of "voltage stability weight 0.6 + power balance weight 0.4" (e.g., 0.6 × 0.99 + 0.4 × 0.9 = 0.954). If the combined index > 0.9, confirm the final optimized output, which includes the corrected distributed energy output, energy storage charging and discharging strategy, and the combined index. For the final optimized output, execute signal distribution to resources at each stage (e.g., local devices, edge nodes, wide-area regions, and supporting facilities), achieving continuous monitoring and response loops under the expanded control range of dynamic adjustment from local to wide-area.

[0145] The signal distribution stage can employ a priority-based distribution algorithm, dividing the optimized output signals into three levels according to their importance:

[0146] Level 1 signals are critical commands related to safety and stability (such as emergency frequency adjustment and voltage over-limit control), and the end-to-end processing delay is controlled within 50ms.

[0147] Secondary signals are routine scheduling instructions (such as power smoothing adjustment and energy storage charging and discharging plans), with delays controlled within 100ms.

[0148] Level 3 signals are for optimizing reference information (such as uploading prediction data and reporting status), with latency controlled within 200ms.

[0149] The resource load rate of each scheduling node is calculated in real time (e.g., the current typical load rate is 75%). Combined with the load balancing algorithm, the signal is distributed to the node with a load rate of less than 80%, ensuring that the overall resource utilization rate is improved by at least 10%. The distribution results are logged for subsequent performance analysis and strategy optimization.

[0150] In the dynamic expansion of the local-wide collaborative control range, relying on a distributed control system, the local control radius is set at approximately 5km, and the wide-area collaborative coverage range is approximately 50km. Control parameters are dynamically adjusted based on network status (e.g., bandwidth allocation ratio adjusted from the initial 0.6 to 0.8), and combined with real-time network latency and transmission bandwidth (e.g., measured equivalent bandwidth of 1.2Gbps), the signal transmission time is calculated to be no more than 20ms, ensuring no significant lag in control commands. Simultaneously, the packet loss rate of the transmission link is continuously monitored (target value <0.01%), and if this limit is exceeded, the system automatically switches to a backup communication channel to maintain the reliability of the control link.

[0151] The continuous monitoring and response cycle collects key operational data at fixed intervals (e.g., every minute) through status monitoring devices deployed at the source, grid, load, and storage stages. This data includes bus voltage, system frequency, active / reactive power, and energy storage state of charge (SOC). Key thresholds are set (e.g., allowable frequency deviation ±0.2Hz, node voltage deviation ≤5%). When monitored values ​​exceed these limits (e.g., frequency drops to 49.8Hz), corresponding adjustment commands are automatically triggered. Combined with historical operational data (e.g., the average frequency over the past 24 hours is 50.02Hz), trend predictions for the next stage of operation are made. If the probability of a key indicator deviating from its target exceeds 70%, the scheduling strategy is adjusted in advance or a margin for adjustment is reserved, forming a closed-loop response mechanism of "monitoring—assessment—decision-execution."

[0152] The above-mentioned links achieve information exchange through a unified data sharing platform. For example, the signal distribution logs are correlated and analyzed with the operation status monitoring data to dynamically optimize resource scheduling strategies, ultimately reducing the overall response time to 85% of the original level, thereby building a complete technical closed loop from signal processing and cross-domain collaborative adjustment to continuous monitoring and optimization.

[0153] This invention first constructs a multi-objective optimization model based on spatiotemporal attributes (time dynamic weights + spatial clustering constraints), and integrates real-time operational data to generate a spatiotemporally adapted initial collaborative scheduling scheme. Then, through full-process interactive analysis and bidirectional adjustment capability assessment, it iteratively optimizes resource allocation ratios using linear programming and genetic algorithms, and dynamically corrects signal thresholds and edge response plans by combining delay-sensitive threshold calibration and compensation mechanisms (signal rearrangement, feedback cycle shortening). Building upon this, it innovatively integrates the autonomous response mechanism with extended fields of the device communication protocol to generate an autonomous response plan supported by the protocol. Furthermore, through wide-area resource collaborative parameter correction (cross-regional delay analysis, deviation threshold expansion), it establishes a collaborative link between local response and wide-area optimization. Finally, through aggregated evaluation of status indicators (voltage stability + power balance) and a closed loop of "signal distribution - continuous monitoring - dynamic adjustment," it achieves optimized output.

Claims

1. A source-grid-load-storage coordinated interactive optimization control method, characterized in that, Includes the following steps: Step S1: Construct a multi-objective optimization model, integrate relevant parameters of distributed energy output, flexible load regulation capability, energy storage system charging and discharging strategy and overall grid operation status to obtain an initial collaborative scheduling scheme; Step S2: Based on the initial coordinated scheduling scheme, iterative calculations are performed through the entire energy interaction process to determine the final resource allocation ratio under the two-way mutual adjustment mechanism between the power grid and distributed energy. Step S3: Generate control signals based on the final resource allocation ratio, and obtain an edge distribution response plan containing the control release signal sequence through real-time feedback from local devices and global load balancing optimization of edge nodes; Step S4: Evaluate the communication delay level and real-time feedback response deviation of the control and release signal sequence, integrate the delay compensation mechanism, and determine the adjustment threshold of the delay-sensitive signal; Step S5: Integrate the delay-sensitive signal adjustment threshold into the correction process of the wide-area resource collaboration parameters to obtain the wide-area resource interaction correction parameters and signal release adjustment scheme; Step S6: Based on the wide area resource interaction correction parameters and signal release adjustment scheme, integrate the autonomous adjustment parameter update path of the autonomous response mechanism with the message triggering and response interface of the device communication protocol, and determine the response coordination scheme under the embedded protocol extension field configuration. Step S7: Use a response coordination scheme to process the local resource allocation execution unit in the edge distribution response plan, generate an adjustment sequence of the integrated device communication protocol delay management rule set, and obtain the autonomous response plan supported by the protocol. Step S8: Combine the wide-area resource interaction correction parameters and generate the final interaction scheme based on the autonomous response plan; Step S9: Obtain the status index data in the final interactive solution, aggregate and evaluate the overall effect of distributed energy output and energy storage system charging and discharging, obtain the final optimized output, and execute signal distribution to resources in each link.

2. The source-grid-load-storage coordinated interactive optimization control method according to claim 1, characterized in that, In step S1, the initial collaborative scheduling scheme is optimized by incorporating dimensional expansion of time and space attributes.

3. The source-grid-load-storage coordinated interactive optimization control method according to claim 1, characterized in that, The process of step S2 includes: From the initial coordinated scheduling scheme, analyze and extract the energy interaction data and initial resource allocation ratio of each stage; Perform dynamic analysis of energy interaction across the entire process, calculate the interaction intensity and regulation demand in layers, and evaluate the grid regulation capacity and distributed energy response capacity in conjunction with the two-way mutual regulation mechanism between the grid and distributed energy. The initial resource allocation ratio is iteratively processed, and the grid regulation capacity and distributed energy response capacity are balanced through algorithm optimization. An optimized allocation scheme that meets the requirements of energy interaction balance in the whole process is output, verified, and the final resource allocation ratio is output.

4. The source-grid-load-storage coordinated interactive optimization control method according to claim 1, characterized in that, The process of step S3 includes: Extract core variables from the final resource allocation ratio, assign weights, and generate control signals through time series analysis algorithms; The control signal is transmitted to the local device to perform status judgment and frequency adjustment, and then the local device is driven to perform real-time feedback and return real-time feedback data. Based on real-time feedback data, the correspondence between signal strength and resource adjustment amount is analyzed to generate autonomous response execution rules and output response mapping paths; Based on the response mapping path, the load balancing of the edge distribution scheme is evaluated and verified, and then the final edge distribution response plan containing the control release signal sequence is generated.

5. The source-grid-load-storage coordinated interactive optimization control method according to claim 1, characterized in that, The process of step S4 includes: For the control and release of signal sequences, collect signal transmission data, and statistically analyze delay data and delay level classification; The preset delay compensation mechanism is invoked based on the delay level classification to adjust the signal delivery sequence; Collect real-time feedback response data, perform deviation threshold judgment, and optimize the feedback cycle; Key time nodes are extracted from the optimized feedback cycle. Combined with the delay-sensitive characteristics, a preliminary signal is output to adjust the threshold range and calibrate. Finally, a delay-sensitive signal is output to adjust the threshold.

6. The source-grid-load-storage coordinated interactive optimization control method according to claim 1, characterized in that, The process of step S5 includes: The initial threshold data for signal adjustment is extracted and analyzed from the delay-sensitive signal adjustment threshold. The initial threshold data is compared with the preset standard, and combined with the cross-regional time delay distribution, to obtain the response deviation value or response deviation evaluation data. The response deviation value is the instantaneous feedback deviation calculated based on the cross-regional time delay mean, and the response deviation evaluation data is the response deviation threshold after cross-regional time delay distribution analysis and expansion. Based on the allocation of wide-area resources and the needs of business scenarios, the response deviation value or response deviation evaluation data is corrected to obtain the wide-area resource interaction correction parameters. Based on the wide-area resource interaction correction parameters, the processing flow for adjusting the threshold of delay-sensitive signals and the cross-regional signal synchronization strategy are optimized to obtain a signal release adjustment scheme.

7. The source-grid-load-storage coordinated interactive optimization control method according to claim 1, characterized in that, The process of step S6 is as follows: Acquire the communication interface data stream of the edge device and classify the message trigger time points and content contained therein; Based on the message trigger classification results, the extended field structure of the device communication protocol is parsed, and field-trigger mapping and priority sequence judgment are performed; Based on the priority sequence configured by the field, the adjustment logic of the autonomous response mechanism and the signal release adjustment scheme are integrated to output the initial scheme for path adjustment; The path adjustment initial scheme is used to adjust the interaction configuration of the device communication interface in real time, and the actual interaction data of the device communication interface is collected, compared with the interaction configuration, and the execution strategy is output. Update the interaction configuration of the device communication interface, synchronously write the updated configuration into the protocol extension field, verify it, and output the response coordination scheme embedded in the protocol extension field configuration.

8. The source-grid-load-storage coordinated interactive optimization control method according to claim 1, characterized in that, The process of step S7 is as follows: Acquire real-time status data of edge distributed nodes, determine the priority ranking of key nodes, obtain local resource availability information based on the ranking results, and output a preliminary resource allocation plan. For the initial resource allocation plan, collect the interaction logs of edge devices, integrate the latency management history records, and output a set of rules for latency optimization; The edge device execution is driven by a delay optimization rule set, the execution feedback is collected and analyzed, and a response mechanism configuration scheme is output. Based on the response mechanism configuration scheme, obtain the compatibility data supported by the protocol, and generate an autonomous response plan supported by the protocol after verification.

9. The source-grid-load-storage coordinated interactive optimization control method according to claim 1, characterized in that, The process of step S8 is as follows: Calculate the deviation between the autonomous response plan and the cloud optimization paradigm, analyze the source of the deviation or verify the parameter benchmark, use optimization algorithms to analyze the wide-area resource interaction correction parameters, and output a set of wide-area resource collaboration parameters; By combining the cross-regional energy allocation index in the wide-area resource interaction correction parameters, the corresponding resource coordination parameters in the wide-area resource coordination parameter set are further decomposed, and regional energy allocation schemes are output and verified. By combining regional energy allocation schemes with wide-area resource interaction correction parameters and optimizing them, the final interaction scheme is output.

10. The source-grid-load-storage coordinated interactive optimization control method according to claim 1, characterized in that, The process of step S9 is as follows: State index data is extracted from the final interactive solution, classified, and then a set of classification indicators is output; among them, the state index data includes voltage stability and power balance. The voltage stability and power balance are evaluated; if the preliminary evaluation results show that a certain indicator exceeds the corresponding threshold range, an in-depth analysis is conducted on the output of distributed energy and the charging and discharging strategy of the energy storage system to locate the anomaly and output the basis for parameter adjustment. Based on the parameter adjustment criteria, the energy storage system's charging and discharging strategy is corrected and the distributed energy output allocation is optimized. The operation adjustment configuration is output, and the voltage stability index and power balance are recalculated. The support vector machine algorithm is used to predict the comprehensive performance of the two indicators and calculate the comprehensive index. If the comprehensive index is greater than 0.9, the final optimized output is confirmed, which includes the corrected distributed energy output, energy storage charging and discharging strategy, comprehensive index, and the execution signal is distributed to resources in each stage.

Images