A method and system for operation control of distributed energy storage in power distribution networks

Abstract

This invention provides a method and system for the operation and control of distributed energy storage in power distribution networks, belonging to the field of distributed energy storage technology. The method includes acquiring relevant data and performing big data analysis to generate a global strategy; adjusting the global strategy locally; converting the localized strategy into operation instructions for each energy storage unit in the power distribution network; predicting the overall execution impact; adjusting the impact instruction parameters of the operation instructions to obtain adjustment instruction parameters; collecting multi-source safety status data and combining it with the adjustment instruction parameters for global safety monitoring; determining the global safety level based on the monitoring results and locally fixed rules; triggering corresponding linkage protection mechanisms; generating a response performance evaluation report; and performing feedback optimization on the big data analysis model and localized adjustment rules to form a closed-loop operation optimization control method. This method ensures global optimality while achieving rapid local response and proactive safety, solving the problem of isolated safety protection.

Description

Technical Field

[0001] This invention relates to the field of distributed energy storage technology, and in particular to a method and system for controlling the operation of distributed energy storage in power distribution networks. Background Technology

[0002] The global energy system is transitioning towards cleaner and lower-carbon energy. The large-scale grid connection of intermittent and highly volatile renewable energy sources such as wind and solar power poses significant challenges to the stability, regulation, and flexibility of the power grid. To address these challenges, energy storage, especially distributed energy storage, is considered a key solution. Its applications have expanded from simple peak-valley arbitrage to multiple fields such as peak and frequency regulation, reactive power support, emergency backup power, and delaying transmission and distribution upgrades. However, current technologies suffer from problems such as high latency in centralized control, unintelligent edge control, passive and isolated security protection, rigid control strategies, and a lack of adaptability.

[0003] Therefore, the present invention provides a method and system for operation control of distributed energy storage in power distribution networks. Summary of the Invention

[0004] This invention provides a method and system for the operation and control of distributed energy storage in power distribution networks. Based on big data analysis, a global optimization strategy is generated. Next, an edge controller performs localized strategy adjustments and command pre-verification. Real-time security assessment is conducted through multi-source data fusion to determine the security level. Based on the level, tiered linkage protection is triggered, and the model and rules are continuously optimized through post-evaluation feedback, forming a self-evolving intelligent control closed loop. This ensures global optimality while achieving rapid local response, solving the problem of high latency in centralized control, transforming passive protection into proactive security, and addressing the problem of isolated security protection. The system can self-adjust according to the operating environment, resolving strategy rigidity.

[0005] This invention provides a method for operation control of distributed energy storage in a power distribution network, comprising:

[0006] Step 1: Obtain relevant data from energy storage units and grid connection points at multiple points in the distribution network, input the relevant data into the big data analysis model for big data analysis, and generate a global strategy based on the analysis results;

[0007] Step 2: Localize the global strategy, convert the localized strategy into operation instructions for each energy storage unit in the distribution network, predict the overall execution impact of the operation instructions, and adjust the impact instruction parameters of the operation instructions to obtain the adjustment instruction parameters.

[0008] Step 3: Collect multi-source security status data, perform global security monitoring based on the adjustment command parameters and the multi-source security status data, and evaluate the current global security status in real time according to the monitoring results and local fixed rules, and determine the global security level.

[0009] Step 4: Trigger the corresponding linkage protection mechanism according to the global state security level, determine the hierarchical response control command and execute it, conduct post-evaluation of the executed command, generate a response performance evaluation report, and perform feedback optimization on the big data analysis model and local fixed rules to form a closed-loop operation optimization control method.

[0010] This invention provides a method for operating and controlling distributed energy storage in a power distribution network. It involves acquiring relevant data from energy storage units and grid connection points at multiple points within the power distribution network, inputting this data into a big data analysis model for analysis, and generating a global strategy based on the analysis results. The method includes:

[0011] Data is acquired through multiple geographically dispersed energy storage units and grid connection points deployed in various locations via edge smart gateways and communication networks.

[0012] Based on the relevant data, a relevant curve is plotted, curve features are marked from the relevant curve, and relevant features are determined based on the curve features;

[0013] Based on the relevant characteristics, the relevant data is input into the big data analysis model for big data analysis to obtain upper-level results, middle-level results, and lower-level results.

[0014] Based on the upper-level results, global optimization conditions are obtained; the middle-level results yield system resource allocation schemes; and the lower-level results provide device-level operating conditions. These are then combined to arrive at a global strategy.

[0015] This invention provides a method for operating and controlling distributed energy storage in a power distribution network. Based on the upper-level results, global optimization conditions are obtained; based on the middle-level results, a system resource allocation scheme is obtained; and based on the lower-level results, device-level operating conditions are obtained. A comprehensive global strategy is then derived, including:

[0016] A basic model is constructed using global optimization conditions as the core input and device-level operating conditions as constraints.

[0017] The system resource allocation scheme is transformed into decision variables and collaborative constraints of a pre-defined optimization model, resulting in an enhanced optimization model;

[0018] The optimization solver is invoked to solve the basic model and the enhanced optimization model, and the optimal numerical solution is output. The optimal numerical solution is then processed in an engineering manner to obtain the global strategy.

[0019] This invention provides a method for operating and controlling distributed energy storage in a distribution network. It involves localizing a global strategy, converting the localized strategy into operation instructions for each energy storage unit in the distribution network, predicting the overall execution impact of these operation instructions, and adjusting the impact instruction parameters to obtain adjusted instruction parameters. The method includes:

[0020] The system receives a global policy and simultaneously collects local multi-source data in real time. Based on the local multi-source data, it determines the local real-time status, compares the global policy with the local real-time status, performs a feasibility check, and determines fine-tuning steps based on the feasibility check results to fine-tune the global policy, thereby deriving a localized policy.

[0021] The localization strategy is parsed, and the operation instructions for each time period in the localization strategy are decomposed according to the predetermined resource allocation scheme and the parsing results to obtain the decomposed instructions.

[0022] A safety sandbox is created based on the local real-time status. The decomposition instructions are simulated and executed in the corresponding decomposition units in the safety sandbox. The impact of the decomposition execution of the decomposition instructions is predicted based on the simulation execution results. The overall execution impact is obtained by combining all the decomposition execution impacts. The decomposition units correspond one-to-one with the energy storage units.

[0023] Based on the impact of decomposition execution, the sensitive instruction parameters in the decomposed instructions are adjusted first, and based on the overall execution impact and the impact of the first adjustment result on the operation instructions, the instruction parameters are adjusted second, resulting in the adjusted instruction parameters.

[0024] This invention provides a method for operation control of distributed energy storage in a distribution network. It involves a first adjustment to sensitive command parameters in the decomposed execution instructions based on the impact of the decomposed execution, and a second adjustment to the command parameters based on the overall execution impact and the impact of the first adjustment on the operation instructions, resulting in adjusted command parameters. The method includes:

[0025] Based on the sensitive instruction parameters that trigger security warnings by decomposing and executing the impact, the allowable adjustment range of the sensitive instruction parameters is calculated, and the sensitive instruction parameters are subjected to amplitude limiting processing within the allowable adjustment range to generate a first adjustment result;

[0026] Conflict analysis is performed on all the first adjustment results. Based on the conflict analysis results, coordination directions and corresponding coordination strategies are formulated to form a coordination-adjustment table.

[0027] The second adjustment is performed on the operation command parameters based on the coordination-adjustment table to obtain the adjusted command parameters.

[0028] This invention provides a method for operating and controlling distributed energy storage in a power distribution network. The method collects multi-source safety status data, performs global safety monitoring based on adjustment command parameters and the multi-source safety status data, and conducts real-time assessment of the current global safety status according to the monitoring results and locally fixed rules, determining the global safety level. The method includes:

[0029] Collect multi-source security status data during the operation of the distribution network, input the multi-source security status data into the preset local fixed rules and security prediction model for global security monitoring, and generate an independent list of security events based on the monitoring results;

[0030] The security incident list is input into a multi-dimensional fusion evaluation algorithm determined by local fixed rules to evaluate the current global security status in real time.

[0031] The global state security level is determined based on the real-time assessment results and the preset security level matrix.

[0032] This invention provides a method for operating and controlling distributed energy storage in a power distribution network. It triggers corresponding linkage protection mechanisms based on the global state security level, determines and executes tiered response control commands, performs post-evaluation of executed commands, generates a response performance evaluation report, and optimizes the big data analysis model and locally fixed rules through feedback, forming a closed-loop operation optimization control method. The method includes:

[0033] Receive the global state security level, query the preset linkage protection mechanism strategy table, and generate a hierarchical response control instruction set corresponding to the global state security level;

[0034] The hierarchical response control command is issued and executed, and high-frequency data acquisition is initiated to comprehensively record the command-response full-link data and form a command-response full-process data packet.

[0035] Based on the instruction-response full-process data packet, a post-evaluation of the linkage response of the executed instruction is performed to generate a response performance evaluation report;

[0036] The response performance evaluation report and the instruction-response full-process data package are synchronized to the cloud. The first feedback to the big data analysis model is determined based on the instruction-response full-process data package. The big data analysis model is then updated and retrained online based on the first feedback. The second feedback to the local solidified rules is determined based on the local performance conclusions in the response performance evaluation report. The local solidified rules are then optimized based on the second feedback, forming a closed-loop operation optimization control method.

[0037] This invention provides a distributed energy storage operation and control system for power distribution networks, comprising:

[0038] Global strategy module: Obtain relevant data from energy storage units and grid connection points at multiple points in the distribution network, input the relevant data into the big data analysis model for big data analysis, and generate global strategies based on the analysis results;

[0039] Localization module: performs localized adjustments to the global strategy, transforms the localized strategy into operation instructions for each energy storage unit in the distribution network, predicts the overall execution impact of the operation instructions, and adjusts the impact instruction parameters of the operation instructions to obtain the adjustment instruction parameters;

[0040] Global monitoring module: Collects multi-source security status data, performs global security monitoring based on adjustment command parameters and the multi-source security status data, and evaluates the current global security status in real time based on the monitoring results and local fixed rules, and determines the global security level.

[0041] Feedback optimization module: Based on the global state security level, it triggers the corresponding linkage protection mechanism, determines and executes hierarchical response control commands, performs post-evaluation on the executed commands, generates a response performance evaluation report, and performs feedback optimization on the big data analysis model and local fixed rules to form a closed-loop operation optimization control method.

[0042] Compared with existing technologies, the beneficial effects of this application are as follows: A global optimization strategy is generated based on big data analysis; secondly, the edge controller performs localized adjustment of the strategy and pre-verification of instructions; real-time security assessment is conducted through multi-source data fusion to determine the security level; hierarchical linkage protection is triggered according to the level; and the model and rules are continuously optimized through post-evaluation feedback, forming a self-evolving intelligent control closed loop. This ensures global optimality while achieving rapid local response, solving the problem of high latency in centralized control, transforming passive protection into proactive security, solving the problem of isolated security protection, enabling the system to self-adjust with the operating environment, and resolving strategy rigidity.

[0043] Other features and advantages of the invention will be set forth in the following description, and will be apparent in part from the description, or may be learned by practicing the invention. The objects and other advantages of the invention may be realized and obtained by means of the structures particularly pointed out in the written description and the accompanying drawings.

[0044] The technical solution of the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. Attached Figure Description

[0045] The accompanying drawings are provided to further illustrate the invention and form part of the specification. They are used in conjunction with embodiments of the invention to explain the invention and do not constitute a limitation thereof. In the drawings:

[0046] Figure 1 This is a flowchart illustrating a distributed energy storage operation control method for a power distribution network provided in an embodiment of the present invention.

[0047] Figure 2 This is a schematic diagram of the structure of a distributed energy storage operation and control system for a power distribution network provided in an embodiment of the present invention. Detailed Implementation

[0048] The preferred embodiments of the present invention will be described below with reference to the accompanying drawings. It should be understood that the preferred embodiments described herein are for illustration and explanation only and are not intended to limit the present invention. Example 1:

[0049] This invention provides a method for controlling the operation of distributed energy storage in a power distribution network, such as... Figure 1 As shown, it includes:

[0050] Step 1: Obtain relevant data from energy storage units and grid connection points at multiple points in the distribution network, input the relevant data into the big data analysis model for big data analysis, and generate a global strategy based on the analysis results;

[0051] Step 2: Localize the global strategy, convert the localized strategy into operation instructions for each energy storage unit in the distribution network, predict the overall execution impact of the operation instructions, and adjust the impact instruction parameters of the operation instructions to obtain the adjustment instruction parameters.

[0052] Step 3: Collect multi-source security status data, perform global security monitoring based on the adjustment command parameters and the multi-source security status data, and evaluate the current global security status in real time according to the monitoring results and local fixed rules, and determine the global security level.

[0053] Step 4: Trigger the corresponding linkage protection mechanism according to the global state security level, determine the hierarchical response control command and execute it, conduct post-evaluation of the executed command, generate a response performance evaluation report, and perform feedback optimization on the big data analysis model and local fixed rules to form a closed-loop operation optimization control method.

[0054] In this embodiment, the node energy storage unit and the grid connection point are the distributed energy storage system and its connection points with the power distribution network, which are the physical objects for data acquisition. For example, a containerized energy storage node in a commercial area is connected to the grid connection point of the power distribution room.

[0055] In this embodiment, the relevant data are real-time operating parameters and status variables collected from the energy storage unit and grid connection point. These include battery SOC, charge / discharge power, grid connection point voltage, and temperature.

[0056] In this embodiment, the big data analysis model is an analysis model built using machine learning and artificial intelligence algorithms. Its input is relevant data, and its output includes upper-level results, middle-level results, and lower-level results. For example, the big data analysis model takes into account real-time SOC, power, electricity price, and other data, and outputs hierarchical results through deep learning algorithms: the upper level generates a peak-valley arbitrage priority strategy; the middle level allocates tasks such as discharging 200kW at station A and charging 100kW at station B; and the lower level provides early warning of abnormal cell voltage differences in cabinet 3 requiring maintenance.

[0057] In this embodiment, the global strategy is a sequence of executable control instructions generated after comprehensively considering global optimization conditions and device-level operating conditions, guiding the coordinated operation of the entire distributed energy storage system. For example, it generates detailed timing control instructions for 10:00-12:00: Energy storage unit 1 charges at 100kW, and energy storage unit 2 discharges at 80kW; 14:00-16:00: All energy storage units discharge at maximum capacity.

[0058] In this embodiment, the localization strategy is an executable solution adapted to local conditions by fine-tuning the global strategy. For example, the original global strategy of discharging at 200kW from 14:00 to 16:00 is adjusted to discharging at 180kW from 14:00 to 15:30, and forced air cooling is activated, which ensures safety while being as close as possible to the global objective.

[0059] In this embodiment, by converting the global policy into a local policy and a dynamic verification mechanism of the local state, combined with the security sandbox for instruction pre-execution and impact prediction, and using a two-stage parameter adjustment algorithm, the local fine-grained adaptation and security optimization of the global policy are achieved, thereby improving the operational safety of distributed energy storage and realizing the adaptive fusion of the global policy and the local real-time state.

[0060] In this embodiment, the overall execution impact is the system-level predicted effect obtained by comprehensively superimposing the simulation impacts of all decomposed instructions. For example, after considering the impacts of each energy storage unit, the predicted total discharge will reach 270kWh, the average system temperature will rise by 6°C, and the total revenue is expected to increase.

[0061] In this embodiment, the parameters affecting the command are those adjustable parameters that primarily affect system performance and economy and are within safety constraints. Examples include optimization parameters such as charge / discharge timing, power change rate, and operating mode switching timing.

[0062] In this embodiment, the adjusted command parameters are a safe and optimized set of command parameters obtained after two levels of adjustment. For example, it is finally determined that energy storage unit No. 1 will discharge at 80kW for 70 minutes, using a ramp start-stop method, with a power change rate of 5kW / minute, achieving the expected benefits, and keeping the temperature rise within a safe range.

[0063] In this embodiment, multi-source safety status data refers to diverse safety-related data collected from various aspects of the power distribution network, including equipment operating status, environmental parameters, and real-time grid information. For example, battery voltage, current, temperature, and SOC are collected from energy storage units; power, frequency, and voltage deviation are collected from grid connection points; ambient temperature, humidity, and smoke concentration are collected from environmental sensors; and real-time safety parameters such as circuit breaker status and fault recording data are collected from protection devices.

[0064] In this embodiment, the locally fixed rules are a set of rules that exist in the form of condition-action, based on expert knowledge and safety judgment logic pre-stored in the local control system. For example, judgment rules based on industry standards and operational experience are triggered if the battery temperature exceeds 45°C, or if the SOC is below 20% and the discharge power is greater than 100kW.

[0065] In this embodiment, global safety monitoring is a comprehensive and systematic process of monitoring and analyzing the overall safety status using locally established rules and safety prediction models. For example, it simultaneously monitors multiple safety indicators such as the thermal runaway risk of all energy storage units, grid connection voltage stability, and environmental fire hazards to form an overall safety situation awareness.

[0066] In this embodiment, the monitoring results are preliminary safety assessments and anomaly identifications derived from global safety monitoring. For example, records of discrete safety anomalies include: abnormal temperature of battery cluster 2 in energy storage cabinet 3 (currently 48°C), excessive voltage fluctuations at grid connection point 5, and communication interruptions of sensors in the fire protection system.

[0067] In this embodiment, the real-time assessment is a comprehensive analysis of the security event list based on a multi-dimensional fusion assessment algorithm to quantify the current global security risk level. For example, using a weighted scoring method, the current comprehensive risk value is calculated to be 75 points, of which abnormal temperature contributes 40 points, voltage fluctuation contributes 25 points, and communication failure contributes 10 points, resulting in an assessment conclusion of a high-risk state.

[0068] In this embodiment, the global state safety level is a comprehensive quantitative indicator used to characterize the overall safety status of the distributed energy storage system. It is typically divided into 1-4 levels, corresponding to normal, early warning, alarm, and dangerous states, respectively, providing a basis for decision-making in hierarchical linkage control. For example, when a battery cluster's temperature reaches 48°C (exceeding the limit) and its insulation resistance decreases but protection is not triggered, the comprehensive risk assessment score is 65 points, and according to the level matrix, it is determined to be level 3, an alarm.

[0069] In this embodiment, the linkage protection mechanism strategy table is a pre-set mapping table between security levels and specific protection actions, defining the standardized response procedures to be executed under different security levels. For example, when the security level is level 3 alarm, the strategy table stipulates that the system should be derated to 80% power, additional cooling should be activated, and the status should be reported every 5 minutes; when the level is level 4 danger, the system stipulates that the PCS contactor should be immediately disconnected, the fire protection system should be activated, and an emergency alarm should be sent.

[0070] In this embodiment, the graded response control instruction set is a standardized, hierarchical set of control commands generated according to the safety level, containing operation instructions of different urgency levels. For example, the instruction set generated for level 3 includes a series of instructions such as limiting PCS power to 80% of the rated value, starting BMS equalization management, switching the air conditioning system to forced cooling mode, and increasing the data acquisition frequency of the monitoring system to 1 time / second.

[0071] In this embodiment, the response performance evaluation report is an evaluation conclusion formed after quantitative analysis of the command execution effect, including performance indicators and problem analysis. For example, the report shows that the power reduction command response delay is 200ms, exceeding the expectation by 100ms, the temperature control effect is good, the temperature drops by 8°C within 5 minutes, and the power overshoot is 5% due to communication delay, among other detailed evaluation results.

[0072] In this embodiment, the system receives the security level and queries the linkage strategy table to generate a hierarchical instruction set, executes the instructions and collects full-link data, generates a performance report through post-evaluation, and synchronizes the analysis model and local rules to the cloud and edge respectively to form a closed-loop optimization and a two-way optimization mechanism, which significantly improves the autonomous optimization capability to cope with abnormal states.

[0073] The working principle and beneficial effects of the above technical solution are as follows: A global optimization strategy is generated based on big data analysis. Next, the edge controller performs localized adjustment of the strategy and pre-verification of instructions. Real-time security assessment is conducted through multi-source data fusion to determine the security level. Based on the level, tiered linkage protection is triggered. Post-evaluation feedback continuously optimizes the model and rules, forming a self-evolving intelligent control closed loop. This ensures global optimality while achieving rapid local response, solving the problem of high latency in centralized control, transforming passive protection into proactive security, resolving the problem of isolated security protection, enabling the system to self-adjust according to the operating environment, and solving the problem of strategy rigidity. Example 2:

[0074] This invention provides a method for operating and controlling distributed energy storage in a power distribution network. The method involves acquiring relevant data from energy storage units and grid connection points at multiple points within the power distribution network, inputting this data into a big data analysis model for analysis, and generating a global strategy based on the analysis results. The method includes:

[0075] Data is obtained from multiple geographically dispersed energy storage units and grid connection points in the power distribution network through edge smart gateways and communication networks deployed in various locations.

[0076] Based on the relevant data, a relevant curve is plotted, curve features are marked from the relevant curve, and relevant features are determined based on the curve features;

[0077] Based on the relevant characteristics, the relevant data is input into the big data analysis model for big data analysis to obtain upper-level results, middle-level results, and lower-level results.

[0078] Based on the upper-level results, global optimization conditions are obtained; the middle-level results yield system resource allocation schemes; and the lower-level results provide device-level operating conditions. These are then combined to arrive at a global strategy.

[0079] In this embodiment, the edge smart gateway and communication network are intelligent data acquisition and communication devices deployed locally at the energy storage station. They support multi-protocol conversion and interact with the cloud via wired or wireless networks. For example, an industrial-grade gateway is equipped with 4G / 5G and Ethernet interfaces to collect data from the energy storage station in real time and upload it to the cloud platform.

[0080] In this embodiment, the relevant curve is a trend chart drawn from the collected relevant data using visualization technology, used to intuitively show the pattern and characteristics of parameter changes over time. For example, a daily load curve is drawn based on a week's load data of an energy storage station. The curve shows that the peak electricity consumption period is from 10:00 to 14:00 each day, with a maximum load of 800kW, while the low electricity consumption period is from 2:00 to 5:00 in the morning, with a load of only about 200kW.

[0081] In this embodiment, curve features are key feature points and segments identified and marked from relevant curves, used to characterize the typical operating states and changing patterns of the system. For example, daily peak load points, valley load points, load ramp-up phases, and load decline phases are marked from the load curve.

[0082] In this embodiment, the relevant features are higher-order characteristic parameters derived from the curve features through further calculation. These parameters can reflect the deep operational characteristics and performance indicators of the system. For example, based on the load curve features, key indicators characterizing the system's operational characteristics, such as daily load factor, peak-to-valley difference rate, and daily average load change rate, are calculated.

[0083] In this embodiment, the upper-level results are macro-level strategy guidance and system-level operational objectives derived from big data analysis models, providing overall guidelines for the operation of the entire distributed energy storage system. For example, the analysis might derive a strategic decision for tomorrow: to prioritize peak-valley arbitrage, fully discharging during the midday peak electricity price period and charging during the nighttime off-peak electricity price period.

[0084] In this embodiment, the intermediate-level results, guided by the upper-level results, are resource coordination and task allocation schemes that clarify the specific operational roles and power allocations of each energy storage unit. For example, a specific implementation plan is formulated, in which energy storage station A is responsible for smoothing the morning load peak with a maximum discharge power of 300kW, and energy storage station B is responsible for supporting the evening load with a maximum discharge power of 200kW.

[0085] In this embodiment, the lower-level results are evaluations of the status of each energy storage unit, reflecting the real-time operating capabilities and physical limitations of the equipment, and providing boundary constraints for control strategies. For example, the evaluation shows that the maximum allowable discharge current of energy storage unit 3 has decreased from 200A to 150A due to cell aging; and that the cooling system efficiency of energy storage unit 5 has decreased, requiring a 10% derating operation, among other equipment-level status information.

[0086] In this embodiment, the global optimization conditions are mathematical optimization objectives and constraints formed by synthesizing the results from the upper layers, which guide the solution direction of the optimization algorithm. For example, an optimization model is established with the objective function of maximizing overall benefits and with constraints such as the SOC constraints, power constraints, and grid interaction power limits of each energy storage unit.

[0087] In this embodiment, the device-level operating conditions are the specific operating boundaries and safety limitations of each energy storage device determined based on the results from the lower layer, ensuring that operating commands are within the safe operating range of the device. For example, specific device constraints are specified, such as the SOC operating range of energy storage unit 1 being 20%-95%, the maximum charging and discharging power being 150kW, and the temperature control range being 15-40℃.

[0088] The working principle and beneficial effects of the above technical solution are as follows: through multi-source data acquisition and feature extraction, a hierarchical analysis architecture is adopted, including upper-layer global optimization, middle-layer resource allocation, and lower-layer device constraints, and finally a collaboratively optimized global control strategy is generated, which takes into account both the global optimization of the system and local operation constraints, realizes the collaborative perception of data across the entire domain, and eliminates information silos. Example 3:

[0089] This invention provides a method for operating and controlling distributed energy storage in a power distribution network. Based on the upper-level results, global optimization conditions are obtained; based on the middle-level results, a system resource allocation scheme is obtained; and based on the lower-level results, device-level operating conditions are obtained. A comprehensive global strategy is then derived, including:

[0090] A basic model is constructed using global optimization conditions as the core input and device-level operating conditions as constraints.

[0091] The system resource allocation scheme is transformed into decision variables and collaborative constraints of a pre-defined optimization model, resulting in an enhanced optimization model;

[0092] The optimization solver is invoked to solve the basic model and the enhanced optimization model, and the optimal numerical solution is output. The optimal numerical solution is then processed in an engineering manner to obtain the global strategy.

[0093] In this embodiment, the core inputs refer to the global optimization objective and economic requirements, which are guiding conditions for model solving. For example, with the core objective of maximizing peak-valley arbitrage profits, the daily profit must not be less than the preset profit, while also meeting the grid's peak-shaving needs by providing at least 1000kW of discharge power during the evening peak hours.

[0094] In this embodiment, constraints refer to the physical limitations and safe operating boundaries of the equipment itself, which are hard conditions that the model must adhere to. For example, the SOC of each energy storage unit must be maintained between 20% and 90%, the charging and discharging power must not exceed the rated power, and the cell temperature must be controlled within the range of 15-40℃.

[0095] In this embodiment, the basic model is a preliminary mathematical model with the optimization objective as the direction and device constraints as the boundary. Its inputs are the optimization objective and device limitations, and the output is the theoretical optimal solution. For example, the objective function is constructed as max(profit) = Σ(discharge profit - charging cost), and the constraints include SOC range, power limit, etc., and the output is the ideal charging and discharging power of each unit.

[0096] In this embodiment, the decision variables are the unknowns that need to be solved in the optimization model, representing controllable operating parameters. For example, the charging and discharging power of each energy storage unit in each time period is set as the decision variable P_j(t), where j is the unit number, t is the time period, and there are a total of 24×q variables, where q is the number of units.

[0097] In this embodiment, the coordination constraint is a set of related restrictions that ensure coordinated operation among different energy storage units. For example, it requires that the total discharge power of all energy storage units must not exceed the grid's capacity limit, and that power changes of each unit must be synchronized to avoid grid impact.

[0098] In this embodiment, the enhanced optimization model is a complete model built upon the basic model, incorporating system coordination requirements and resource allocation schemes. Its inputs include coordination constraints in addition to basic conditions, and its output is a more realistic and executable solution. For example, by adding constraints such as regional total power balance and mobile energy storage scheduling timing to the basic model, a practically applicable optimization model can be formed.

[0099] In this embodiment, the optimization solver is a specialized computational software tool used to solve mathematical optimization problems. For example, using CPLEX or Gurobi solvers, the objective function and constraints of the enhanced optimization model are input, and the optimal solution is calculated using algorithms such as the simplex method and the interior point method.

[0100] In this embodiment, the optimal numerical solution is the theoretically optimal result output by the solver, usually in numerical form. For example, the solver outputs the specific power values ​​of each energy storage unit at 96 time points in the next 24 hours, such as P1(t1)=50kW, P1(t2)=-30kW, with negative values ​​indicating charging, etc.

[0101] In this embodiment, the engineering process is the process of transforming the theoretical numerical solution into a control strategy that can be safely executed. For example, the power curve output by the solver is smoothed to avoid sudden power changes; start-stop delay protection is added; and the instructions are checked to see if they exceed the real-time capabilities of the device, ultimately generating a set of operational instructions that can be issued.

[0102] In this embodiment, the global strategy is the final executable plan derived after engineering processing. For example: 10:00-12:00: Energy storage unit 1 charges at 100kW; 14:00-16:00: All units discharge at maximum capacity, with a total power not exceeding 2000kW, is a detailed set of timing control instructions.

[0103] The working principle and beneficial effects of the above technical solution are as follows: A two-layer modeling architecture is adopted, combining global optimization conditions with device-level operating conditions into a basic model. The overall resource allocation scheme is transformed into decision variables and collaborative constraints to form an enhanced optimization model. Global strategies are generated through optimization solutions and engineering processing. This achieves a deep integration of theoretical optimization and practical engineering constraints, improving overall operational efficiency. Example 4:

[0104] This invention provides a method for operating and controlling distributed energy storage in a distribution network. It involves localizing a global strategy, converting the localized strategy into operation instructions for each energy storage unit in the distribution network, predicting the overall execution impact of these operation instructions, and adjusting the impact instruction parameters to obtain adjusted instruction parameters. The method includes:

[0105] The system receives a global policy and simultaneously collects local multi-source data in real time. Based on the local multi-source data, it determines the local real-time status, compares the global policy with the local real-time status, performs a feasibility check, and determines fine-tuning steps based on the feasibility check results to fine-tune the global policy, thereby deriving a localized policy.

[0106] The localization strategy is parsed, and the operation instructions for each time period in the localization strategy are decomposed according to the predetermined resource allocation scheme and the parsing results to obtain the decomposed instructions.

[0107] A safety sandbox is created based on the local real-time status. The decomposition instructions are simulated and executed in the corresponding decomposition units in the safety sandbox. The impact of the decomposition execution of the decomposition instructions is predicted based on the simulation execution results. The overall execution impact is obtained by combining all the decomposition execution impacts. The decomposition units correspond one-to-one with the energy storage units.

[0108] Based on the impact of decomposition execution, the sensitive instruction parameters in the decomposed instructions are adjusted first, and based on the overall execution impact and the impact of the first adjustment result on the operation instructions, the instruction parameters are adjusted second, resulting in the adjusted instruction parameters.

[0109] In this embodiment, local multi-source data refers to diverse data collected in real time from various sensors and devices at the energy storage site, including operating parameters, environmental data, and equipment status. For example, battery pack SOC, voltage, and cell temperature are obtained from the BMS; charging and discharging power and efficiency are obtained from the PCS; cabin temperature and humidity are obtained from environmental sensors; and real-time data such as grid connection point power factor and grid frequency are obtained from smart meters.

[0110] In this embodiment, the local real-time status is an assessment of the system's current operating status based on a comprehensive analysis of local multi-source data, reflecting the device's immediate capabilities and limitations. For example, the comprehensive analysis might conclude that the current system's safe discharge power limit is 180kW, but due to the high temperature of battery cluster 3, it needs to be derated; the cooling system is working normally, but the continuously rising ambient temperature requires close monitoring.

[0111] In this embodiment, feasibility verification involves comparing and analyzing the issued global policy with the local real-time state to verify whether the policy can be safely executed locally. For example, the verification might find that the global policy requires a discharge of 200kW, but the local maximum allowable discharge power is only 180kW, and the ambient temperature has reached the warning value, thus determining that the instruction is not feasible under the current conditions.

[0112] In this embodiment, the fine-tuning step is a strategy adjustment method and procedure determined based on the feasibility verification results, used to adapt the global strategy to the local actual situation. For example, specific adjustment measures such as formulating a proportional reduction in power command by 10%, advancing the discharge time by 30 minutes, and activating the backup cooling unit are implemented.

[0113] In this embodiment, the parsing result is the instruction composition and timing relationship obtained after structural analysis of the localization strategy. For example, the parsed strategy contains three main instructions: power control instruction, duration instruction, and cooling control instruction, as well as the execution timing and dependencies of each instruction.

[0114] In this embodiment, the decomposition instruction breaks down the composite instruction in the localization strategy into specific operation commands for each energy storage unit. For example, the total discharge power of 180kW is decomposed into: 100kW for energy storage unit 1 and 80kW for energy storage unit 2, and corresponding PCS control instructions are generated for each.

[0115] In this embodiment, the security sandbox is a virtual simulation environment built based on local real-time status, used to test the execution effect of instructions without affecting the actual system. For example, a digital twin environment containing the current battery status, temperature parameters, and power grid conditions is created to rehearse the instruction execution process.

[0116] In this embodiment, the decomposition unit simulation execution involves virtually executing the decomposition instructions for each energy storage unit within a secure sandbox. For example, in a digital twin environment, energy storage unit 1 is simulated to discharge at 100kW for 90 minutes, and its impact on battery temperature, SOC change, and lifespan loss is calculated in real time.

[0117] In this embodiment, the decomposition execution impact refers to the predicted effects and impact assessments of the commands of a single energy storage unit after simulation execution. For example, simulation shows that discharging at 100kW for 90 minutes by decomposition unit 1 will cause the cell temperature to rise by 8°C, the SOC to decrease by 25%, and the lifespan loss to increase by 0.01%.

[0118] In this embodiment, each decomposition unit corresponds one-to-one with an energy storage unit, and each virtual unit in the simulation test establishes a completely corresponding mapping relationship with the actual physical energy storage unit. For example, decomposition unit 1 in the safety sandbox completely replicates the real-time state of the actual energy storage unit 1, ensuring that the simulation results are realistic and reliable.

[0119] In this embodiment, sensitive command parameters are those critical command parameters that have a significant impact on system security and require priority adjustment. Examples include core parameters that directly affect device safety, such as charging / discharging power values, runtime, and start / stop frequency.

[0120] In this embodiment, the first adjustment is a preliminary safety adjustment to sensitive instruction parameters based on the decomposition of execution effects. For example, if the simulation shows an excessive temperature rise, the discharge power is reduced from 100kW to 80kW, and the duration is shortened from 90 minutes to 70 minutes.

[0121] In this embodiment, the second adjustment is a fine-grained optimization of the influencing command parameters based on the overall execution impact, while ensuring safety. For example, adjusting the power change rate to make it smoother, optimizing the charging and discharging timing to better match electricity price changes, and improving economic benefits.

[0122] The working principle and beneficial effects of the above technical solution are as follows: By using a global policy-local state dynamic verification mechanism, combined with a security sandbox for instruction pre-execution and impact prediction, and employing a two-stage parameter adjustment algorithm, the global policy can be localized for fine-grained adaptation and security optimization, thereby improving the operational safety of distributed energy storage and achieving adaptive fusion of global policy and local real-time state. Example 5:

[0123] This invention provides a method for operating and controlling distributed energy storage in a distribution network. The method involves a first adjustment to sensitive command parameters in a decomposed command based on the impact of the decomposed execution, and a second adjustment to command parameters based on the overall execution impact and the first adjustment result, resulting in adjusted command parameters. The method includes:

[0124] Based on the sensitive instruction parameters that trigger security warnings by decomposing and executing the impact, the allowable adjustment range of the sensitive instruction parameters is calculated, and the sensitive instruction parameters are subjected to amplitude limiting processing within the allowable adjustment range to generate a first adjustment result;

[0125] Conflict analysis is performed on all the first adjustment results. Based on the conflict analysis results, coordination directions and corresponding coordination strategies are formulated to form a coordination-adjustment table.

[0126] The second adjustment is performed on the operation command parameters based on the coordination-adjustment table to obtain the adjusted command parameters.

[0127] In this embodiment, the allowable adjustment range refers to the numerical range within which sensitive command parameters can be adjusted while ensuring system safety. This range is determined by both the physical limits of the device and its real-time status. For example, for the discharge power parameter, the current allowable adjustment range is calculated as [0, 150kW], where 0 is the lower limit, i.e., the non-discharge state, and 150kW is the safe upper limit determined based on battery temperature, SOC, and health status. Exceeding this range may lead to overheating or over-discharge risks.

[0128] In this embodiment, the limiting process is a procedure that forcibly restricts sensitive command parameters within the allowable adjustment range, ensuring that the parameter values ​​are always within the safe boundary. For example, when a command requests a discharge of 180kW, if the system detects that the value exceeds the allowable range [0, 150kW], it will automatically limit it to 150kW and record the adjustment magnitude as a reduction of 30kW and the adjustment reason as a temperature warning.

[0129] In this embodiment, the first adjustment result is a preliminary adjustment scheme generated after limiting the various sensitive command parameters, including parameter adjustment values ​​and adjustment basis. For example, the generated first adjustment result includes: limiting the discharge power of energy storage unit 1 from 200kW to 150kW, shortening the operating time of energy storage unit 2 from 120 minutes to 90 minutes, and including an adjustment explanation due to the temperature approaching the threshold.

[0130] In this embodiment, conflict analysis is a process of checking whether there are contradictions or mutual influences among all the first adjustment results, ensuring that the adjusted instructions can be executed in a coordinated manner. For example, the analysis may find that after the power of energy storage unit 1 is reduced, energy storage unit 2 needs to take on additional load, but its remaining capacity is insufficient, resulting in a conflict between power demand and insufficient capacity; or a resource contention conflict caused by multiple units simultaneously applying to use limited cooling resources.

[0131] In this embodiment, the coordination direction and corresponding coordination strategy are the solution direction and specific methods determined by the conflict analysis, aiming to eliminate conflicts and optimize overall performance. For example, for power allocation conflicts, the coordination direction is to reallocate power tasks, and the strategy is to activate the reserve capacity of energy storage unit No. 3 to share the 20kW load; for resource contention conflicts, the coordination direction is to use resources during off-peak hours, and the strategy is for units to start their cooling equipment in different time periods.

[0132] In this embodiment, the coordination-adjustment table is a structured adjustment guide generated by the system, which clarifies the coordination strategies and specific adjustment methods for various conflicts, and is used to guide the second adjustment. For example, the coordination-adjustment table includes entries such as: Conflict type: uneven power distribution; Coordination strategy: activate standby unit; Adjustment method: adjust the power of energy storage unit No. 4 from 0 to 50kW, forming a systematic set of adjustment instructions.

[0133] In this embodiment, the second adjustment is a fine-tuning of the influencing command parameters based on the coordination-adjustment table, which optimizes the overall system performance while eliminating conflicts. For example, according to the coordination-adjustment table, the power of energy storage unit No. 4 is increased to 50kW to share the load, while the power change rate of energy storage unit No. 1 is adjusted to 10kW / minute to achieve a smooth transition, and the operating sequence of each unit is optimized to avoid resource conflicts.

[0134] In this embodiment, the adjustment command parameters are a set of safe, coordinated, and optimized command parameters obtained after two levels of adjustment. For example, the final adjustment result is: energy storage unit 1 discharges 150kW for 80 minutes, energy storage unit 2 discharges 100kW for 70 minutes, and energy storage unit 4 discharges 50kW for 60 minutes. All energy storage units adopt a time-sharing start-up strategy, and the total system discharge power is 300kW, which fully meets the safety constraints and operational requirements.

[0135] The working principle and beneficial effects of the above technical solution are as follows: Based on the decomposition and execution of sensitive command parameters, the allowable adjustment range is calculated and the amplitude is limited to generate the first adjustment result. Through conflict analysis, a coordination strategy is formulated to form a coordination-adjustment table. Based on this, the second adjustment is carried out to obtain the adjustment command parameters, thereby achieving accurate identification and rapid handling of safety warnings and improving the accuracy and reliability of safety control. Example 6:

[0136] This invention provides a method for controlling the operation of distributed energy storage in a power distribution network. The method collects multi-source safety status data, performs global safety monitoring based on adjustment command parameters and the multi-source safety status data, and conducts real-time assessment of the current global safety status according to the monitoring results and locally fixed rules, determining the global safety level. The method includes:

[0137] Collect multi-source security status data during the operation of the distribution network, input the multi-source security status data into the preset local fixed rules and security prediction model for global security monitoring, and generate an independent list of security events based on the monitoring results;

[0138] The security incident list is input into a multi-dimensional fusion evaluation algorithm determined by local fixed rules to evaluate the current global security status in real time.

[0139] The global state security level is determined based on the real-time assessment results and the preset security level matrix.

[0140] In this embodiment, R represents the global risk value; The time decay factor of the i-th event; The overall score of the i-th event; β represents the weighting coefficient of the most severe event; Represents the highest overall score among all events; γ represents the risk dispersion weighting coefficient; This represents the standard deviation of the overall score for all events. This represents the average score of all events; i represents the index of the security event; and n represents the total number of security events.

[0141] In this embodiment, the safety prediction model is a predictive analysis model built based on machine learning algorithms, capable of predicting short-term safety trends based on the current state. Its input is multi-source safety state data, and its output is the safety risk prediction result. For example, inputting the current battery temperature, cooling system status, and load condition, the output is the temperature change curve and overheat probability for the next 30 minutes; inputting the current insulation resistance value and humidity, the output predicts the future insulation fault risk level.

[0142] In this embodiment, the independent security event list is a structured list of events formed by organizing the monitoring results. Each event includes attributes such as type, location, and severity. For example, the list includes entries such as Event 1: Temperature Anomaly - Cabinet 3, Cluster 2 - Level 2, Event 2: Voltage Fluctuation - Site 5 - Level 1, and Event 3: Communication Failure - Fire Protection System - Level 3, along with timestamps and detailed parameters.

[0143] In this embodiment, the safety level matrix is ​​a predefined standard for classifying safety levels, mapping quantified risk assessment results to specific safety levels. For example, the matrix specifies: 0-30 points: Level 1 Normal, 31-60 points: Level 2 Warning, 61-80 points: Level 3 Alarm, 81-100 points: Level 4 Danger. The current assessment score is 75 points, corresponding to Level 3 Alarm status.

[0144] The working principle and beneficial effects of the above technical solution are as follows: collect multi-source security status data, generate a security event list through local fixed rules and security prediction models for global security monitoring, conduct real-time evaluation using a multi-dimensional fusion evaluation algorithm, and finally determine the global security level based on the security level matrix. Example 7:

[0145] This invention provides a method for operating and controlling distributed energy storage in a power distribution network. Based on the global state security level, it triggers corresponding linkage protection mechanisms, determines and executes tiered response control commands, performs post-evaluation on the executed commands, generates a response performance evaluation report, and optimizes the big data analysis model and locally fixed rules through feedback, forming a closed-loop operation optimization control method. The method includes:

[0146] Receive the global state security level, query the preset linkage protection mechanism strategy table, and generate a hierarchical response control instruction set corresponding to the global state security level;

[0147] The hierarchical response control command is issued and executed, and high-frequency data acquisition is initiated to comprehensively record the command-response full-link data and form a command-response full-process data packet.

[0148] Based on the instruction-response full-process data packet, a post-evaluation of the linkage response of the executed instruction is performed to generate a response performance evaluation report;

[0149] The response performance evaluation report and the instruction-response full-process data package are synchronized to the cloud. The first feedback to the big data analysis model is determined based on the instruction-response full-process data package. The big data analysis model is then updated and retrained online based on the first feedback. The second feedback to the local solidified rules is determined based on the local performance conclusions in the response performance evaluation report. The local solidified rules are then optimized based on the second feedback, forming a closed-loop operation optimization control method.

[0150] In this embodiment, high-frequency data refers to fine-grained operational data collected during instruction execution, used to accurately capture the system's dynamic response process. For example, during the execution of a power reduction instruction, changes in battery voltage, current, and temperature are collected at a frequency of 100Hz, and the power adjustment process is recorded every 10ms to capture transient response characteristics.

[0151] In this embodiment, the command-response end-to-end data records all relevant data from the issuance of the command to the completion of the device's execution, forming a complete operation traceability chain. For example, this includes data from all stages such as the command issuance timestamp, PCS reception confirmation signal, actual power change curve, device status feedback, and protection device action records.

[0152] In this embodiment, the instruction-response end-to-end data packet is a structured data set formed by time-aligning and encapsulating the end-to-end data, used for subsequent analysis. For example, the data packet contains instruction set metadata, high-frequency acquired electrical parameters, equipment status change sequences, environmental parameter records, etc., and is accompanied by a unified time synchronization identifier.

[0153] In this embodiment, the local performance conclusion is an evaluation result of the response effect of a specific device or subsystem, reflecting the local performance. For example: PCS 1 responds slowly, with an instruction execution delay of 300ms; the cooling system performance is reduced, with a temperature drop rate 20% lower than the standard value; the BMS equalization function is effective, with a voltage deviation reduction of 0.1V, etc.

[0154] In this embodiment, coordinated response refers to the overall performance of multiple system components acting in concert according to control commands. For example, upon receiving a Level 3 command, the PCS performs power reduction, the BMS initiates equalization, the cooling system enhances heat dissipation, and the monitoring system increases the sampling rate, among other subsystems working together in a coordinated response process.

[0155] In this embodiment, the first feedback refers to optimization suggestions for the cloud-based big data analysis model, proposed based on pattern features in the entire process data packet. For example, historical data shows that the temperature prediction model has a large deviation under high-temperature conditions, suggesting the addition of training samples for sudden temperature changes; there is a systematic delay between power control commands and actual responses, requiring correction of timing parameters and other model optimization inputs.

[0156] In this embodiment, the second feedback is an adjustment suggestion for the locally fixed rules, proposed based on local performance conclusions. For example, for the response delay problem of PCS 1, it is suggested to add a command timeout resend mechanism; for insufficient cooling system performance, it is suggested to modify the fan start-up threshold from 40℃ to 35℃, and other rule optimization suggestions.

[0157] The working principle and beneficial effects of the above technical solution are as follows: it receives the security level and queries the linkage strategy table to generate a hierarchical instruction set, executes the instructions and collects full-link data, generates a performance report through post-evaluation, and synchronizes the analysis model and local rules to the cloud and edge respectively to form a closed-loop optimization, forming a two-way optimization mechanism, which significantly improves the autonomous optimization capability to cope with abnormal states. Example 8:

[0158] This invention provides a distributed energy storage operation and control system for power distribution networks, such as... Figure 2 As shown, it includes:

[0159] Global strategy module: Obtain relevant data from energy storage units and grid connection points at multiple points in the distribution network, input the relevant data into the big data analysis model for big data analysis, and generate global strategies based on the analysis results;

[0160] Localization module: performs localized adjustments to the global strategy, transforms the localized strategy into operation instructions for each energy storage unit in the distribution network, predicts the overall execution impact of the operation instructions, and adjusts the impact instruction parameters of the operation instructions to obtain the adjustment instruction parameters;

[0161] Global monitoring module: Collects multi-source security status data, performs global security monitoring based on adjustment command parameters and the multi-source security status data, and evaluates the current global security status in real time based on the monitoring results and local fixed rules, and determines the global security level.

[0162] Feedback optimization module: Based on the global state security level, it triggers the corresponding linkage protection mechanism, determines and executes hierarchical response control commands, performs post-evaluation on the executed commands, generates a response performance evaluation report, and performs feedback optimization on the big data analysis model and local fixed rules to form a closed-loop operation optimization control method.

[0163] The working principle and beneficial effects of the above technical solution are as follows: A global optimization strategy is generated based on big data analysis. Next, the edge controller performs localized adjustment of the strategy and pre-verification of instructions. Real-time security assessment is conducted through multi-source data fusion to determine the security level. Based on the level, tiered linkage protection is triggered. Post-evaluation feedback continuously optimizes the model and rules, forming a self-evolving intelligent control closed loop. This ensures global optimality while achieving rapid local response, solving the problem of high latency in centralized control, transforming passive protection into proactive security, resolving the problem of isolated security protection, enabling the system to self-adjust according to the operating environment, and solving the problem of strategy rigidity.

[0164] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for operation and control of distributed energy storage in a power distribution network, characterized in that, include: Step 1: Obtain relevant data from energy storage units and grid connection points at multiple points in the distribution network, input the relevant data into the big data analysis model for big data analysis, and generate a global strategy based on the analysis results; Step 2: Localize the global strategy, convert the localized strategy into operation instructions for each energy storage unit in the distribution network, predict the overall execution impact of the operation instructions, and adjust the impact instruction parameters of the operation instructions to obtain the adjustment instruction parameters. Step 3: Collect multi-source security status data, perform global security monitoring based on the adjustment command parameters and the multi-source security status data, and evaluate the current global security status in real time according to the monitoring results and local fixed rules, and determine the global security level. Step 4: Trigger the corresponding linkage protection mechanism according to the global state security level, determine the hierarchical response control command and execute it, conduct post-evaluation of the executed command, generate a response performance evaluation report, and perform feedback optimization on the big data analysis model and local fixed rules to form a closed-loop operation optimization control method.

2. The method for operation and control of distributed energy storage in a distribution network according to claim 1, characterized in that, Relevant data is obtained from energy storage units and grid connection points at multiple points in the distribution network. This data is then input into a big data analysis model for analysis. Based on the analysis results, a global strategy is generated, including: Data is acquired through multiple geographically dispersed energy storage units and grid connection points deployed in various locations via edge smart gateways and communication networks. Based on the relevant data, a relevant curve is plotted, curve features are marked from the relevant curve, and relevant features are determined based on the curve features; Based on the relevant characteristics, the relevant data is input into the big data analysis model for big data analysis to obtain upper-level results, middle-level results, and lower-level results. Based on the upper-level results, global optimization conditions are obtained; the middle-level results yield system resource allocation schemes; and the lower-level results provide device-level operating conditions. These are then combined to arrive at a global strategy.

3. The method for operation and control of distributed energy storage in a distribution network according to claim 2, characterized in that, Based on the upper-level results, global optimization conditions are obtained; the middle-level results yield a system resource allocation scheme; and the lower-level results provide device-level operating conditions. These are then combined to derive a global strategy, including: A basic model is constructed using global optimization conditions as the core input and device-level operating conditions as constraints. The system resource allocation scheme is transformed into decision variables and collaborative constraints of a pre-defined optimization model, resulting in an enhanced optimization model; The optimization solver is invoked to solve the basic model and the enhanced optimization model, and the optimal numerical solution is output. The optimal numerical solution is then processed in an engineering manner to obtain the global strategy.

4. The method for operation and control of distributed energy storage in a distribution network according to claim 1, characterized in that, The global strategy is locally adjusted, transforming it into operational instructions for each energy storage unit in the distribution network. The overall execution impact of these operational instructions is predicted, and the impact instruction parameters are adjusted to derive the adjusted instruction parameters, including: The system receives a global policy and simultaneously collects local multi-source data in real time. Based on the local multi-source data, it determines the local real-time status, compares the global policy with the local real-time status, performs a feasibility check, and determines fine-tuning steps based on the feasibility check results to fine-tune the global policy, thereby deriving a localized policy. The localization strategy is analyzed, and the operation instructions for each time period in the localization strategy are decomposed according to the predetermined resource allocation scheme and the analysis results to obtain the decomposed instructions. A safety sandbox is created based on the local real-time status. The decomposition instructions are simulated and executed in the corresponding decomposition units in the safety sandbox. The impact of the decomposition execution of the decomposition instructions is predicted based on the simulation execution results. The overall execution impact is obtained by combining all the decomposition execution impacts. The decomposition units correspond one-to-one with the energy storage units. Based on the impact of decomposition execution, the sensitive instruction parameters in the decomposed instructions are adjusted first, and based on the overall execution impact and the impact of the first adjustment result on the operation instructions, the instruction parameters are adjusted second, resulting in the adjusted instruction parameters.

5. The method for operation and control of distributed energy storage in a distribution network according to claim 4, characterized in that, Based on the impact of decomposition execution, a first adjustment is made to the sensitive instruction parameters in the decomposed instructions. Based on the overall execution impact and the impact of the first adjustment on the operation instructions, a second adjustment is made to the instruction parameters, resulting in adjusted instruction parameters, including: Based on the sensitive instruction parameters that trigger security warnings by decomposing and executing the impact, the allowable adjustment range of the sensitive instruction parameters is calculated, and the sensitive instruction parameters are subjected to amplitude limiting processing within the allowable adjustment range to generate a first adjustment result; Conflict analysis is performed on all the first adjustment results. Based on the conflict analysis results, coordination directions and corresponding coordination strategies are formulated to form a coordination-adjustment table. The second adjustment is performed on the operation command parameters based on the coordination-adjustment table to obtain the adjusted command parameters.

6. The method for operation and control of distributed energy storage in a distribution network according to claim 1, characterized in that, Collect multi-source security status data, perform global security monitoring based on adjustment command parameters and the multi-source security status data, and conduct real-time assessment of the current global security status based on the monitoring results and locally fixed rules, and determine the global security level, including: Collect multi-source security status data during the operation of the distribution network, input the multi-source security status data into the preset local fixed rules and security prediction model for global security monitoring, and generate an independent list of security events based on the monitoring results; The security incident list is input into a multi-dimensional fusion evaluation algorithm determined by local fixed rules to evaluate the current global security status in real time. The global state security level is determined based on the real-time assessment results and the preset security level matrix.

7. The method for operation and control of distributed energy storage in a distribution network according to claim 1, characterized in that, Based on the global state security level, corresponding linkage protection mechanisms are triggered, hierarchical response control commands are determined and executed, and the executed commands are post-evaluated to generate a response performance evaluation report. Furthermore, feedback optimization is performed on the big data analysis model and locally fixed rules to form a closed-loop operation optimization control method, including: Receive the global state security level, query the preset linkage protection mechanism strategy table, and generate a hierarchical response control instruction set corresponding to the global state security level; The hierarchical response control command is issued and executed, and high-frequency data acquisition is initiated to comprehensively record the command-response full-link data and form a command-response full-process data packet. Based on the instruction-response full-process data packet, a post-evaluation of the linkage response of the executed instruction is performed to generate a response performance evaluation report; The response performance evaluation report and the instruction-response full-process data package are synchronized to the cloud. The first feedback to the big data analysis model is determined based on the instruction-response full-process data package. The big data analysis model is then updated and retrained online based on the first feedback. The second feedback to the local solidified rules is determined based on the local performance conclusions in the response performance evaluation report. The local solidified rules are then optimized based on the second feedback, forming a closed-loop operation optimization control method.

8. A distributed energy storage operation and control system for power distribution networks, characterized in that, include: Global strategy module: Obtain relevant data from energy storage units and grid connection points at multiple points in the distribution network, input the relevant data into the big data analysis model for big data analysis, and generate global strategies based on the analysis results; Localization module: performs localized adjustments to the global strategy, transforms the localized strategy into operation instructions for each energy storage unit in the distribution network, predicts the overall execution impact of the operation instructions, and adjusts the impact instruction parameters of the operation instructions to obtain the adjustment instruction parameters; Global monitoring module: Collects multi-source security status data, performs global security monitoring based on adjustment command parameters and the multi-source security status data, and evaluates the current global security status in real time based on the monitoring results and local fixed rules, and determines the global security level. Feedback optimization module: Based on the global state security level, it triggers the corresponding linkage protection mechanism, determines and executes hierarchical response control commands, performs post-evaluation on the executed commands, generates a response performance evaluation report, and performs feedback optimization on the big data analysis model and local fixed rules to form a closed-loop operation optimization control method.

Images