A park energy storage and load collaborative scheduling method, system, device and medium of triple nested logic

Abstract

This invention discloses a method, system, equipment, and medium for coordinated scheduling of energy storage and load in a park using a triple-nested logic, belonging to the field of smart grid and energy storage operation control technology. The method includes: acquiring park operation data and generating first scheduling prediction information; generating executable power candidate schemes based on the prediction information and real-time operation data; performing action frequency constraint processing, establishing an action recovery mechanism, and verifying it; implementing cooling window control, limiting the execution time, power change amplitude, and sequence of actions; and then performing comprehensive judgment to generate the final scheduling command, and adaptively correcting the scheduling parameters based on execution feedback. This invention, through triple-nested logic, can limit the number of start-stop cycles, direction switching, and power ramp-up cycles in real time under conditions of high-proportion renewable energy access and rapid load fluctuations, actively plan cooling windows, and optimize action sequences, avoiding thermal instability and voltage surges caused by frequent charging and discharging or continuous high-power operation.

Description

Technical Field

[0001] This invention relates to the field of smart grid and energy storage operation control technology, specifically to a method, system, equipment and medium for coordinated scheduling of park energy storage and load using triple nested logic. Background Technology

[0002] With the rapid integration of distributed photovoltaic, wind power, and other new energy sources into the industrial park's distribution network, the power load exhibits significant fluctuations and uncertainties. To buffer power changes and balance load and power supply, large-capacity energy storage devices are commonly deployed in industrial parks, and energy storage is used in conjunction with load scheduling to improve the absorption rate of new energy sources and the reliability of power supply.

[0003] However, existing energy storage and load dispatching methods are mostly based on single forecasting and power control models. Their control logic and feedback mechanisms are relatively simple, making it difficult to meet the safety and economic operation requirements under high-proportion renewable energy access. In the dispatching forecasting phase, conventional methods usually only perform a power or temperature forecast once before dispatching begins and use static thresholds for constraints. They lack continuous tracking and dynamic verification of future trends, making it difficult for dispatching schemes to remain accurate and stable when external loads or meteorological conditions change abruptly.

[0004] In the operation and control phase, existing solutions often only limit the maximum charging and discharging power or state of charge range of the energy storage device, failing to comprehensively constrain it from multiple dimensions such as operation frequency, continuous time interval and power ramp-up speed. This leads to the energy storage device frequently starting and stopping or repeatedly charging and discharging in a short period of time, which can easily cause safety hazards such as excessive temperature rise and excessive mechanical stress.

[0005] In terms of temperature management, traditional practices mostly adopt passive control, reducing power or shutting down only when the temperature reaches the upper limit. They do not combine temperature rise prediction with equipment cooling capacity in the scheduling and planning stage, calculate the necessary cooling time window in advance, and lack a systematic arrangement for the execution sequence of multiple power change actions. As a result, they cannot effectively prevent heat accumulation and lifespan degradation caused by continuous high-power operation.

[0006] Existing solutions have limited feedback and correction capabilities during execution, and lack an automated adaptive adjustment mechanism for deviations between predictions and actual conditions. Once external conditions change rapidly, they can only rely on manual intervention or periodic rescheduling, making it difficult to achieve closed-loop management of prediction, control, and feedback. Summary of the Invention

[0007] In view of the above-mentioned problems, the present invention is proposed.

[0008] Therefore, the technical problem solved by this invention is: how to ensure that energy storage devices in a power system containing large-scale distributed new energy and energy storage devices maintain temperature safety, stable state of charge, and stable grid-connected power while meeting external dispatch requirements, under conditions of uncertain power forecasting, frequent load fluctuations, and heat accumulation caused by long-term continuous operation of equipment, so as to achieve safe, efficient, and long-term stable operation of the power system in the park.

[0009] To solve the above-mentioned technical problems, the present invention provides the following technical solution: a method for coordinated scheduling of park energy storage and load with triple nested logic, which includes acquiring the operating data of park energy storage devices and load equipment, and generating first scheduling prediction information;

[0010] Based on predictive information and real-time operational data, executable power candidate schemes are generated according to multi-dimensional band rules.

[0011] Perform action frequency constraint processing on executable power candidate schemes, establish an action recovery mechanism, and verify it based on the operating status;

[0012] Cooling window control is implemented for the validated executable power candidate schemes to limit the execution time, power change magnitude and sequence of actions;

[0013] The executable power candidate schemes processed by the cooling window are comprehensively evaluated to generate the final energy storage and load coordinated scheduling instructions, and the scheduling parameters are adaptively corrected based on the execution feedback.

[0014] As a preferred embodiment of the triple-nested logic-based park energy storage and load coordinated scheduling method described in this invention, the step of acquiring the operating data of the park energy storage device and load equipment, and generating the first scheduling prediction information includes,

[0015] Voltage, current, temperature and active power acquisition units are installed at each battery cluster, grid connection interface and load branch of the energy storage device in the park. Instantaneous voltage, current, state of charge, charge and discharge rate, equipment temperature rise and load power operation data are collected through the data acquisition terminal at a set time window.

[0016] The collected data is timestamped using a unified clock, and smoothing is performed on the collected data with the scheduling cycle as the time window to remove random disturbances and obtain a time-continuous running data sequence.

[0017] A short-term trend prediction model is constructed based on the operational data sequence to calculate the future changes of park load power, distributed new energy output, energy storage state of charge, energy storage unit temperature and grid-connected power within the scheduling cycle;

[0018] The calculated load power prediction sequence, new energy output prediction sequence, energy storage state of charge prediction sequence, energy storage temperature rise prediction sequence, and grid-connected power prediction sequence are combined to form the first scheduling prediction information, and the energy storage capacity safety boundary and power change characteristics within the prediction interval are extracted simultaneously.

[0019] As a preferred embodiment of the triple-nested logic-based park energy storage and load coordinated scheduling method described in this invention, the step of generating executable power candidate schemes based on prediction information and real-time operating data according to multi-dimensional zone rules includes:

[0020] The first scheduling prediction information is sealed and not executed directly; instead, the second scheduling information from the outside world is received.

[0021] Using the first scheduling prediction information as a reference, the charging and discharging power range of energy storage is calculated step by step under the conditions of satisfying the energy storage state of charge, temperature rise, ramp rate and grid connection safety boundary;

[0022] Combining the quantitative targets for peak shaving, energy replenishment, and electricity price response in external instructions, the optimal power value is selected in each time step, and the optimal planned power values ​​are sorted by time to form a preliminary candidate power scheme;

[0023] Action recognition is performed on the preliminary candidate power schemes. Based on the recognition results and multi-dimensional band rules, the preliminary candidate power schemes are constrained to generate executable power candidate schemes.

[0024] As a preferred embodiment of the triple-nested logic-based park energy storage and load coordinated scheduling method described in this invention, the step of performing action frequency constraint processing on executable power candidate schemes, establishing an action recovery mechanism, and verifying based on the operating status includes:

[0025] When the start-up, shutdown, direction switching, or power ramping action of the energy storage device is detected, the corresponding credit value is deducted from the credit balance of the corresponding action according to the power amplitude of the action and the temperature zone, state of charge zone, and grid-connected power fluctuation zone.

[0026] Based on the first dispatch prediction information, determine whether the temperature, state of charge and grid-connected power fluctuation of the energy storage device meet the preset limit recovery conditions, and execute the limit recovery or suspend recovery.

[0027] Based on the cooldown time in the first scheduling prediction information, a sliding window is determined. Within the sliding window, the occurrence of various constraint actions is accumulated and statistically analyzed. The statistical results are then compared with the corresponding action quota balance and the upper limit of the window at the current time step to determine the violation time step.

[0028] Corrective operations are performed sequentially based on the violation time step until the corresponding action quota balance and window limit are met; the execution power scheme after correction and a complete action statistics report are output as input information for cooling window control.

[0029] As a preferred embodiment of the triple-nested logic-based park energy storage and load coordinated scheduling method described in this invention, the step of implementing cooling window control on verified executable power candidate schemes, limiting the execution time, power change amplitude, and sequence of actions, includes...

[0030] Based on the temperature prediction sequence in the first scheduling prediction information, combined with the rated cooling coefficient of the energy storage device and the maximum allowable ramp rate, the minimum cooling time required after high-power operation is calculated, and a continuous cooling window is established within the scheduling cycle accordingly.

[0031] If the planned execution time of the power change action in the candidate scheme falls within the cooling window, the execution time and power are adjusted to meet the limitations of the cooling window.

[0032] If the power change rate of the action performed after the cooling window exceeds the rated ramp rate, it is split into several continuous execution segments with single-step change rates that meet the requirements.

[0033] This invention, through predictive-driven cooling window generation and fixed-sequence power execution, ensures the thermal stability of the energy storage device during continuous operation while controlling the frequency of operations. The window length is derived from temperature prediction and quantitative calculation of cooling capacity, allowing cooling intervals to be planned in advance before scheduling.

[0034] By separating the sequential control and ramp rate during execution, the electrical shock and localized thermal accumulation caused by the simultaneous triggering of several high-power actions are prevented. Compared with traditional passive power reduction or single temperature threshold shutdown, this invention can continuously maintain the internal thermal balance of the energy storage unit and ensure stable grid-connected power output.

[0035] As a preferred embodiment of the triple-nested logic-based park energy storage and load coordinated scheduling method described in this invention, the step of implementing cooling window control on verified executable power candidate schemes, limiting the execution time, power change amplitude, and sequence of actions, further includes...

[0036] When a cooling window ends, if several power change actions are triggered simultaneously at the same time, they will be executed sequentially according to a fixed safety order, including:

[0037] First, perform the discharging action, then the charging action, and within the same type of action, perform them in order of decreasing power change.

[0038] When the power change amplitude is the same, the grid-connected side power adjustment is performed first, and then the internal power adjustment of the energy storage device is performed.

[0039] In the final verification before the action is executed, the predicted values ​​of temperature and state of charge in the first scheduling prediction information are retrieved again and compared with the real-time collected temperature and state of charge measurement values. If the real-time measurement shows insufficient cooling, the current cooling window will be automatically extended. If the state of charge has not recovered to the mid-band region, the power will be reduced again until the predicted conditions are consistent with the actual measurement values.

[0040] As a preferred embodiment of the triple-nested logic-based park energy storage and load coordinated scheduling method described in this invention, the step of comprehensively determining the executable power candidate schemes processed by the cooling window, generating the final energy storage and load coordinated scheduling instruction, and adaptively correcting the scheduling parameters based on execution feedback includes:

[0041] After the candidate power schemes have undergone action frequency constraint processing and cooling window control in sequence, the modified power schemes are comprehensively judged and the final energy storage and load coordinated scheduling command is generated.

[0042] The comprehensive judgment includes checking whether the power amplitude, execution time and action sequence are still consistent with the future temperature, state of charge and grid-connected power prediction results of the first scheduling prediction information, and verifying whether they meet the latest second scheduling information;

[0043] During the execution of instructions, the actual charging and discharging power, state of charge and temperature of the energy storage device are continuously collected and compared with the corresponding predicted values ​​in the first scheduling prediction information.

[0044] When any key parameter deviates from the predicted value beyond the set allowable range, the execution order is automatically corrected in the next scheduling time step, and the reason for the correction is recorded.

[0045] After completing a single correction, the final energy storage and load coordinated scheduling command is sent to the field control device, and a complete execution record is generated.

[0046] This invention provides a campus energy storage and load collaborative scheduling system with triple nested logic.

[0047] To solve the above-mentioned technical problems, the present invention provides the following technical solution: a park energy storage and load coordinated scheduling system with triple nested logic, comprising: a predictive scheduling module, a candidate scheme generation module, a verification module, a cooling module, and an output module;

[0048] The predictive scheduling module acquires the operating data of the park's energy storage devices and load equipment, and generates the first scheduling prediction information;

[0049] The candidate scheme generation module generates executable power candidate schemes based on prediction information and real-time operating data, in accordance with multi-dimensional band rules.

[0050] The verification module performs action frequency constraint processing on executable power candidate schemes, establishes an action recovery mechanism, and verifies them based on the operating status.

[0051] The cooling module implements cooling window control on the verified executable power candidate schemes, limiting the execution time, power change amplitude, and sequence of actions;

[0052] The output module comprehensively evaluates the executable power candidate schemes processed by the cooling window, generates the final energy storage and load coordinated scheduling instruction, and adaptively corrects the scheduling parameters based on the execution feedback.

[0053] The present invention provides a computer device, including a memory and a processor, wherein the memory stores a computer program, characterized in that the processor executes the computer program to implement the steps of the above-described triple-nested logic-based campus energy storage and load coordinated scheduling method.

[0054] The present invention provides a computer-readable storage medium having a computer program stored thereon, characterized in that, when the computer program is executed by a processor, it implements the steps of the above-described triple-nested logic-based campus energy storage and load coordinated scheduling method.

[0055] The beneficial effects of this invention are as follows: Based on the first scheduling prediction information, this invention combines the predicted sequences of future load power, new energy output, energy storage state of charge, energy storage temperature rise, and grid-connected power with real-time operating data to dynamically generate a scheduling scheme and simultaneously generate capacity safety boundaries and power change characteristics, achieving direct connection between prediction and execution. By controlling start-up, shutdown, direction switching, and power ramp-up times through action frequency limits and recovery mechanisms, it avoids overheating and voltage surges caused by frequent charging and discharging of energy storage devices. By controlling the execution time, amplitude, and sequence of power changes through a cooling window based on temperature prediction, and by verifying the predicted and measured values ​​in real time before execution, it prevents heat accumulation and lifespan degradation caused by continuous high-load operation. During the execution phase, scheduling parameters are automatically corrected based on real-time deviations to ensure that instructions continuously meet external scheduling requirements and safety boundaries. Attached Figure Description

[0056] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0057] Figure 1 The above is a flowchart of a three-layer nested logic-based campus energy storage and load coordinated scheduling method provided in one embodiment of the present invention.

[0058] Figure 2 The flowchart illustrates the generation of candidate scheduling schemes for energy storage and load using a triple-nested logic-based collaborative scheduling method for park energy storage and load, as provided in one embodiment of the present invention.

[0059] Figure 3 This is a flowchart for verifying the operation status of a triple-nested logic-based park energy storage and load coordinated scheduling method according to an embodiment of the present invention. Detailed Implementation

[0060] To make the above-mentioned objects, features, and advantages of the present invention more apparent and understandable, specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of them. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the protection scope of the present invention.

[0061] Existing energy storage and load coordination technologies in industrial parks typically rely solely on single power prediction and simple threshold control, which have several shortcomings in actual operation:

[0062] Existing technical solutions lack continuous trend calculations and dynamic corrections for future load power, renewable energy output, energy storage state of charge, energy storage temperature rise, and grid-connected power. This makes it difficult for dispatching schemes to respond in a timely manner when external conditions change. Furthermore, they only use power limits or state of charge ranges as constraints, which cannot comprehensively constrain the number of start-stop cycles, direction switching, and power ramp-up. This can easily lead to frequent start-stop cycles and excessive direction switching of energy storage devices in a short period of time, causing local overheating and voltage surges.

[0063] Example 1, referring to Figure 1 This is one embodiment of the present invention, which provides a method for coordinated scheduling of park energy storage and load based on triple nested logic, including:

[0064] S1. Obtain the operating data of the energy storage devices and load equipment in the park, and generate the first scheduling prediction information.

[0065] S2. Based on the predicted information and real-time operation data, generate candidate scheduling schemes for energy storage and load according to the multi-dimensional zone rules.

[0066] S3. Perform action frequency constraint processing on candidate solutions, establish an action recovery mechanism, and verify it based on the running status.

[0067] S4. Implement cooling window control for the verified candidate solutions, limiting the execution time, power change range, and sequence of actions.

[0068] S5. The candidate schemes processed by the cooling window are comprehensively evaluated to generate the final energy storage and load coordinated scheduling instructions, and the scheduling parameters are adaptively corrected based on the execution feedback.

[0069] This invention addresses the problems of existing park energy storage and load scheduling technologies, such as the disconnect between prediction and execution, the single dimension of action constraints, the lack of proactive cooling planning, and the lack of real-time feedback correction, by acquiring and processing multi-source operational data to construct first scheduling prediction information, generating multi-dimensional zone scheduling candidate schemes, establishing action frequency quotas and recovery mechanisms, implementing cooling window control, and executing comprehensive judgment and adaptive correction technology.

[0070] Example 2, refer to Figures 2-3 As an embodiment of the present invention, based on the previous embodiment, a method for coordinated scheduling of park energy storage and load with triple nested logic is provided, including:

[0071] S1. Obtaining the operating data of the park's energy storage devices and load equipment, and generating the first scheduling prediction information includes the following steps:

[0072] S11. Voltage, current, temperature and active power acquisition units are installed at each battery cluster, grid connection interface and load branch of the energy storage device in the park, and instantaneous voltage and current, state of charge, charge and discharge rate, equipment temperature rise and load power operation data are collected through the data acquisition terminal in a set time window.

[0073] The collected data is timestamped using a unified clock, and smoothing is performed on the collected data using the scheduling cycle as the time window to remove random disturbances and obtain a time-continuous running data sequence.

[0074] S12. Based on the operational data sequence, a short-term trend prediction model is constructed to calculate the future changes of park load power, distributed new energy output, energy storage state of charge, energy storage unit temperature and grid-connected power within the scheduling cycle.

[0075] At the start of the scheduling process, the system first collects real-time operating data of the energy storage devices and load equipment in the park, including the voltage, current, state of charge (SoC), instantaneous charge and discharge power, temperature of each energy storage battery cluster, as well as the load power and distributed new energy output parameters of the park. The system then performs unified clock alignment and smoothing on the data to remove random disturbances and obtain a continuous time series.

[0076] S13. Further, the system constructs a short-term trend prediction model to calculate the future changes of park load power, distributed new energy output, energy storage state of charge, energy storage unit temperature and grid-connected power within the scheduling cycle. The calculated load power prediction sequence, new energy output prediction sequence, energy storage state of charge prediction sequence, energy storage temperature rise prediction sequence and grid-connected power prediction sequence are combined to form the first scheduling prediction information, and the energy storage capacity safety boundary and power change characteristics within the prediction interval are extracted simultaneously.

[0077] The formula for calculating the predicted load power is:

[0078] ,

[0079] in, For the first Step load power forecast value, The average load is calculated within the rolling window. The load power correlation coefficient, For the first Step load power prediction value.

[0080] The expression for calculating the predicted output of new energy sources is as follows:

[0081] ,

[0082] in, For the first Step by new energy power output forecast, For the new energy power calculated within the rolling window, For the first Step by new energy power output forecast, Correlation coefficient for contribution to new energy sources.

[0083] It should be further explained that the load power correlation coefficient Correlation coefficient with new energy output This data is obtained by statistically analyzing historical load power and actual renewable energy output data from the dispatching park, and calculating the linear correlation coefficient between data points at adjacent time points. Since power load and renewable energy output are positively correlated... The typical value range is 0.7 to 0.95. The typical value range is 0.5 to 0.9. When deploying the system, the initial value is calculated using historical data from a certain period (e.g., one month); after the system is running, this coefficient can be updated periodically to adapt to changes.

[0084] Load power and renewable energy output together constitute the net demand of the park at every moment in the future, which is a direct constraint on energy storage charging and discharging strategies.

[0085] The predicted state of charge of energy storage is calculated using the following expression:

[0086] ,

[0087] in, For the first Step-by-step energy storage state of charge prediction value, For the first Step-by-step energy storage state of charge prediction value, For rated energy storage capacity, For the sampling time step, For the first Step-by-step energy storage power baseline value, For discharge efficiency, For charging efficiency.

[0088] The formula for calculating the predicted temperature rise of energy storage is as follows:

[0089] ,

[0090] in, For the first Predicted temperature rise in energy storage For the first Predicted temperature rise in energy storage The equivalent heat source coefficient, is the thermal time constant.

[0091] It should be further explained that the equivalent heat source coefficient The thermal time constant is typically set between 0.3 and 1.0, depending on the cell's heat generation rate and cooling conditions. The calibration time is typically 3–15 minutes, based on the actual heat dissipation characteristics of the cooling structure. Both parameters must be determined by combining the manufacturer's thermal test report with on-site measurement data, with a safety margin included.

[0092] The predicted grid-connected power is calculated using the following expression:

[0093] ,

[0094] in, For the first Step-by-step grid-connected power prediction value, For the first Step-by-step energy storage power baseline value.

[0095] It should be further noted that the energy storage power baseline value needs to be initialized when calculating the first scheduling prediction information. ,in To predict the total number of steps.

[0096] The actual measured power value of the energy storage device at the current moment is set and obtained in real time from the data acquisition system.

[0097] For all future moments , uniformly set as The value is based on the most conservative safety assumptions, that is, before optimization calculations are performed, it is assumed that the energy storage device does not operate during the scheduling cycle, the zero power is within the range of the device's allowable power, and is the safest neutral value among them.

[0098] The zero-power assumption at future moments during initialization only provides a starting point for generating preliminary prediction information and will not be used as the final scheduling instruction. In step S2, the system will recalculate the optimal energy storage power sequence based on the current prediction information, combined with external scheduling requirements and security constraints, through an optimization algorithm. This sequence will completely cover the initial zero-power assumption and will then proceed to subsequent processing stages such as action frequency constraints and cooling window control.

[0099] Therefore, the specific initial values ​​do not affect the generation of the final scheduling scheme, as long as they are within the allowable range of the device. During implementation, the future values ​​are first... Initialize to zero, and then obtain the actual scheduling power through optimization.

[0100] S14. Combine the calculated load power prediction sequence, new energy output prediction sequence, energy storage state of charge prediction sequence, energy storage temperature rise prediction sequence and grid-connected power prediction sequence to form the first scheduling prediction information, and simultaneously extract the energy storage capacity safety boundary and power change characteristics within the prediction interval.

[0101] All predictions are based on the same sampling step size. Δ The prediction steps H are generated, with timestamps perfectly aligned. The five types of predictions are then applied at each prediction time step. Combined into data tuples:

[0102] ,

[0103] in, For the current scheduling time, This is the prediction step number.

[0104] Example: For instance, the scheduling cycle of a certain park is 15 minutes, and the sampling step size is 1 minute. The prediction model outputs the sequence of the next 15 steps: the load power gradually increases from 0.9MW to 1.3MW, distributed photovoltaic power remains at 0.4MW, if energy storage does not operate, the SoC decreases from 68% to 66%, the temperature slowly rises from 34℃ to 36℃, and the grid-connected power is between 0.5 and 0.9MW.

[0105] This invention constructs a short-term trend prediction model that can simultaneously calculate the evolution trajectory of external power demand (load, new energy) and key internal states (state of charge SoC, temperature) of energy storage equipment within future scheduling cycles. The generated first scheduling prediction information breaks through the limitations of traditional single power prediction and realizes the integration of energy flow and equipment thermal dynamics in the park system.

[0106] Reference Figure 2 S2. Based on forecast information and real-time operational data, generating candidate scheduling schemes for energy storage and load according to multi-dimensional zone rules includes the following steps:

[0107] S21. The first scheduling prediction information is sealed and not executed directly; the second scheduling information from the outside world is received.

[0108] It should be noted that the system receives scheduling demand information from the dispatch center or the park's energy management system, including quantitative requirements such as the upper limit of grid-connected peak power, allowed reverse power, expected SoC target range, peak reduction target, off-peak energy replenishment target, and cost or revenue indicators for the next scheduling cycle.

[0109] For example, external dispatch requirements include that the grid-connected power should not exceed 1.0MW in the next 15 minutes, at least 0.1MWh of energy should be replenished during off-peak hours, the energy storage SoC should be maintained between 65% and 75%, and the goal is to smooth the load through charging and discharging so that the total power change rate of the park does not exceed 0.2MW / minute.

[0110] It should also be noted that, without modifying the first scheduling prediction information, the feasible charging and discharging power range of energy storage at each future time step is solved step by step based on the scheduling requirements and the rated power, allowable ramp rate, and minimum duration constraints of the energy storage equipment.

[0111] S22. Using the first scheduling prediction information as a reference, calculate the charging and discharging power range that the energy storage can achieve under the conditions of satisfying the energy storage state of charge, temperature rise, ramp rate and grid connection safety boundary according to the time step.

[0112] By combining the quantitative objectives of peak shaving, energy replenishment, and electricity price response in external instructions, the optimal planned power value is selected in each time step, and the optimal planned power values ​​are sorted by time to form candidate power schemes.

[0113] For example, the system makes a planned power allocation decision based on the known rated power value, allowable power change rate, and preset minimum continuous operating time of the energy storage device, while comprehensively considering the initial value and trend of the energy storage state of charge, the maximum allowable operating temperature of the system, and multiple static operating boundaries such as the power limit requirements of the grid connection point. The priority is to suppress load peaks, and the secondary goal is to supplement the energy storage.

[0114] Specifically, the system directly sets explicit planned power values ​​for different time periods. For example, in the initial stage, energy storage units are arranged to remain on standby to monitor the system status. Subsequently, during the load surge period, they are instructed to discharge at the planned power value to smooth out peak loads. Finally, during the load slump period, they are instructed to charge at another planned power value to restore energy storage levels. By arranging the planned power values ​​for all time periods in chronological order, a preliminary planned power value sequence covering the entire scheduling cycle is formed, i.e., a preliminary candidate power scheme.

[0115] S23. Perform action recognition on the preliminary candidate power schemes, and based on the recognition results combined with multi-dimensional band rules, perform constraint processing on the preliminary candidate power schemes to generate executable power candidate schemes.

[0116] The multidimensional zone rule is a decision-making method that evaluates three types of operating parameters in real time: grid-connected power fluctuation, energy storage device temperature, and energy storage state of charge. Based on preset objective thresholds, these parameters are quantified into discrete levels (i.e., "zones"), and the maximum allowable power change of the energy storage device at the current moment is dynamically calculated based on this comprehensive level.

[0117] Further, the planned power of each step in the preliminary candidate power scheme is read sequentially over time and compared with the planned power of the previous time step. A change in planned power from zero to non-zero or vice versa is considered a start / stop action. A change in the sign of the planned power is considered a direction switching action. A change in planned power exceeding the allowable value dynamically calculated from predicted planned power fluctuations and equipment ramp-up capability is considered a ramp-up action.

[0118] It should be noted that the dynamically calculated allowable value is the maximum allowable power change per unit time given in the technical parameters of the energy storage device, which is used as the starting capacity value for power change.

[0119] The system combines the first scheduling prediction information to evaluate and classify the three types of operating states at the current time step:

[0120] First, determine the intensity of fluctuations in grid-connected power, and determine whether the maximum change within a fixed time window is a small, medium, or large fluctuation.

[0121] Second, determine the temperature of the energy storage device to identify whether the current temperature is in a low, medium, or high temperature zone.

[0122] Third, determine the degree of proximity between the energy storage state of charge and the upper and lower limits of capacity, and determine whether the corresponding parameters are in the middle, near or extreme zones.

[0123] It should be further noted that the fluctuation intensity mentioned in this invention is divided into three levels: small fluctuation, medium fluctuation, and large fluctuation. The dividing threshold is determined according to the requirements of stable operation of the power grid and the power change capability of the equipment.

[0124] It should be noted that the thresholds for small and medium fluctuations are usually set with reference to the general requirements of the power grid for power fluctuations during steady-state operation, or based on the equipment's ability to smoothly regulate power under normal operating conditions; the thresholds for large fluctuations are usually set with reference to the technical requirements of the power grid for extreme fluctuations or fault ride-through, or based on the equipment's ability to withstand the maximum instantaneous power surge.

[0125] The temperature of energy storage devices is divided into low-temperature zone, medium-temperature zone, and high-temperature zone. The boundary thresholds are determined based on the optimal operating temperature range of the equipment and the safety alarm temperature.

[0126] It should be noted that the low-band and mid-band boundary thresholds are typically set as the upper limit of the optimal operating temperature range recommended by the equipment manufacturer. Below this value, the equipment is considered to be in a comfortable operating range with high efficiency and low losses. The high-band boundary threshold is set as the warning temperature or derating start temperature defined in the equipment's safe operation manual. Above this value, the equipment is considered to be under significant thermal stress, requiring proactive power or frequency limiting.

[0127] Energy storage state of charge is divided into mid-band, near-band, and extreme-band regions, and the boundary thresholds are determined based on the device's optimal SoC operating range and battery protection thresholds.

[0128] It should be noted that the middle zone should be set to the optimal daily operating range recommended by the battery management system (BMS) or the operation manual (such as 30%~70%), within which charging and discharging has the least impact on battery life.

[0129] The adjacent band is usually set as the primary protection threshold of the BMS (such as 10%~20% and 80%~90%). When the SoC enters this area, the system needs to carefully plan the depth of charge and discharge.

[0130] The limiting band is usually set as the hard protection cutoff threshold of the BMS (such as 0%~10% and 90%~100%), prohibiting any operation that would cause the SoC to further exceed this boundary, in order to ensure battery safety.

[0131] The system further pre-determines an influence coefficient for each level or zone of each state. This coefficient is set based on equipment safety operation experience, historical test data, and safety regulations, with a typical value range between 0.6 and 1.0. The closer the operating state is to the safety boundary (e.g., greater fluctuations, higher temperatures, or a state of charge closer to its limit), the smaller the corresponding influence coefficient value. The system uses a multiplication method to comprehensively calculate the influence coefficients corresponding to the three states to obtain the comprehensive influence coefficient for the current time step.

[0132] Based on all the processing results of steps S21 to S23, an executable power candidate scheme is generated.

[0133] The planned power of the candidate scheme at the current time step is compared with the planned power at the previous time step, and the actual power change is calculated. If the actual change exceeds the allowable range, the current time step is determined to be a ramp-up action, and quota deduction and constraint processing are triggered; if it does not exceed the allowable range, it is not determined to be a ramp-up action.

[0134] It should be noted that the allowable range is calculated by multiplying the equipment's basic capacity by the comprehensive influence coefficient, and then by the time step, which gives the maximum allowable power change at the current time step.

[0135] The current technical solution is not to simply generate a power plan, but to dynamically quantify three key safety parameters—grid-connected power fluctuation, energy storage device temperature, and state of charge—into small / medium / large, low / medium / high, and medium / near / limit discrete state zones based on objective standards of the equipment and the power grid.

[0136] Based on this, the system calculates a dynamic impact coefficient that comprehensively reflects the real-time safety margin for each scheduling moment, thereby accurately and adaptively scaling down the allowable power variation capability of energy storage. This ensures that the generated executable power candidate schemes, while responding to external economic objectives such as peak shaving and valley filling, have embedded preventative safety constraints.

[0137] Furthermore, by comparing power at adjacent moments, the system identifies three types of subsequent physical actions to be controlled: start-stop, direction switching, and hill climbing. Each action is then bound to a multi-dimensional safety status label at that moment, i.e., the zone to which it belongs.

[0138] Therefore, the output of step S2 is an information base that includes safety semantics and action tags. It is not only a power sequence, but also provides a quantifiable decision basis for action frequency constraints and cooling window control.

[0139] Reference Figure 3 S3. Perform action frequency constraint processing on candidate solutions, establish an action recovery mechanism, and verify based on the running status, including the following steps:

[0140] S31. When the start-up, shutdown, direction switching, or power ramping action of the energy storage device is detected, the corresponding credit value is deducted from the credit balance of the corresponding action according to the power amplitude of the action and the temperature zone, state of charge zone, and grid-connected power fluctuation zone.

[0141] It should be noted that a base deduction value is defined for each type of action (start / stop, direction switching, ramping), typically one unit. This base value reflects the standard consumption of the equipment caused by the action. For each dimension (temperature, state of charge, grid-connected power fluctuation), a preset influence coefficient (e.g., ranging from 1.0 to 2.0) is assigned to each zone (e.g., high, medium, low)). The higher the coefficient, the higher the cost or risk of performing the action in that state. The system multiplies the base deduction value by the corresponding multiple zone influence coefficients based on the current action type and the specific zone combination to obtain the final deduction amount. This amount is typically rounded up to a natural number.

[0142] Example: A start-stop operation occurs at the 6th minute (power increases from 0MW to 0.5MW). The system detects that the current temperature and state of charge are both in the middle band, while the grid-connected power fluctuation is in the low band. According to the preset deduction rules, the system determines that the impact coefficient of this operation is 1.0 for each band under this combination of middle / medium / low bands. Therefore, the deduction is 1 (1×1×1×1=1), and it is deducted from the start-stop credit balance.

[0143] At the 8th minute, the system identified a directional switching action (power changed from +1.0MW to -0.3MW) and a ramping action (power change of 1.3MW). At this time, the energy storage device temperature was in the high zone (influence coefficient set to 1.5), the state of charge was in the adjacent zone (influence coefficient set to 1.2), and the grid-connected power fluctuation was in the large fluctuation level (influence coefficient set to 1.3).

[0144] Based on the same set of deduction rules, the system calculates the deduction amount for the direction switching action as follows: the base deduction value of 1 multiplied by each influence coefficient, i.e., 1×1.5×1.2×1.3=2.34, which is rounded up to get 3; similarly, the deduction amount for the hill climbing action is also 3. The system deducts the amount from the remaining credit limit for both the direction switching and hill climbing actions.

[0145] S32. Based on the first dispatch prediction information, determine whether the temperature, state of charge and grid-connected power fluctuation of the energy storage device meet the preset limit recovery conditions, and execute the limit recovery or pause recovery.

[0146] The specific credit recovery conditions are divided into the following three levels: If, according to the first scheduling prediction information, it is determined that the temperature of the energy storage device has dropped to the low zone of the temperature zone and the state of charge is in the middle zone of the state of charge zone, and the change in grid-connected power in the first scheduling prediction information within adjacent time steps does not exceed the power fluctuation threshold preset in the device at the time of deployment, then the corresponding credit will be restored according to the standard recovery rate preset in the device.

[0147] When only one or two of the temperature conditions, state of charge conditions, and grid-connected power fluctuation conditions are met, the quota recovery is performed according to the preset low-rate recovery ratio of the equipment; when none of the three conditions are met, the recovery of the corresponding action quota is suspended.

[0148] It should be noted that the power fluctuation threshold preset by the equipment is set according to the technical specifications of the grid connection point or the grid dispatch requirements, and is used to define the allowable power fluctuation range under normal operation.

[0149] The standard recovery rate reflects the equipment's rated recovery capability under ideal, safe conditions; the low-rate recovery ratio is used to conservatively reduce the recovery speed under non-ideal operating conditions. The specific values ​​for the standard recovery rate and the low-rate recovery ratio can be determined through analysis of historical operating data or based on guidance parameters provided by the equipment manufacturer.

[0150] Example: At the 12th minute, based on the first dispatch prediction information, the system determines that the current predicted temperature of the energy storage device is 34℃, which has reached and is within the low zone of the temperature zone division; the predicted state of charge is 68%, which is within the middle zone of the state of charge zone division; simultaneously, the change in grid-connected power within adjacent time steps is calculated to be 0.1MW, which does not exceed the preset power fluctuation threshold in the equipment parameter table. Since the three conditions of temperature, state of charge, and grid-connected power fluctuation simultaneously meet all the requirements of standard recovery, the system then replenishes the credit balance for start-stop and ramp-up actions once each, according to the preset standard recovery rate in the equipment parameter table.

[0151] S33. Based on the cooldown time in the first scheduling prediction information, determine the sliding window, accumulate statistics on the occurrence of various constraint actions within the sliding window, and compare the statistical results with the corresponding action quota balance and window upper limit value at the current time step to determine the violation time step.

[0152] Further constraints include the number of start-stop actions, the number of direction changes, the total change in ramp power, and the duration during which the actual charging and discharging power of the energy storage device remains in a non-zero state.

[0153] The sliding window length is determined by combining the cooling time of the first scheduling prediction information. Within any continuous window time period, the number of start-stop actions, the number of direction switching actions, the change in total climbing power, and the duration during which the actual charging and discharging power of the energy storage device is continuously in a non-zero state are accumulated. Each accumulated value obtained in the current continuous time interval is compared one by one with the quota balance of the corresponding action calculated in real time at the current scheduling time step and the upper limit of the window allowed at the current time step.

[0154] If the comparison results show that any cumulative value exceeds the quota balance of the corresponding action or exceeds the window limit of the current time step, then the current scheduling time step is directly judged as an illegal time step.

[0155] Example: Suppose the preset statistical window length is 3 minutes. Within the window from the 9th to the 11th minute, a total of 3 start / stop actions occur. The system's preset maximum allowed start / stop actions for this window is 2. Because the cumulative value of 3 exceeds the window's maximum of 2, the system determines the 11th minute, the end of the current statistical window, as a violation.

[0156] S34. Perform correction operations sequentially based on the violation time step until the corresponding action quota balance and window limit are met; output the execution power scheme after correction operation and a complete action statistics report as input information for cooling window control.

[0157] For time steps that are judged to be in violation, power downgrading, action delay and action merging correction operations are performed in sequence, and the violation judgment of S33 is immediately re-executed after each correction until the planned power of the current time step meets all quota and window constraints.

[0158] The planned charging and discharging power is further reduced step by step at the current time step. The power change for each reduction is set to the minimum adjustment range based on the equipment control accuracy. After each reduction, the credit balance and window statistics are re-verified to ensure they meet the requirements, until the power is reduced to zero.

[0159] If the constraints are still not met after the power level is reduced to zero, the entire power change action planned for the current time step will be postponed to the first subsequent time step that is assessed to meet all the allowance and window constraints. The planned power of the original non-compliant time step will be set to zero or the safe power value of the previous moment will be maintained.

[0160] Furthermore, if after the delay there are still several planned actions where the single-step power change amplitude is less than the minimum adjustment amplitude, and the consecutive power change actions cannot meet the constraints, then the power change actions of the current adjacent time step will be merged into one consecutive power change action with a single-step change rate not exceeding the rated ramp rate of the equipment, and the matching with the quota balance and window limit will be verified again until the current time step meets all requirements.

[0161] Example: At the 11th minute, a violation is detected due to insufficient start-stop capacity. The system first attempts to reduce the power level, decreasing the charging power from 0.4MW to 0.2MW in preset steps (e.g., 0.1MW), but this still fails. Next, it attempts to delay the action, moving the charging action to the earliest time (13 minutes) that meets the constraints, while setting the power to zero at the 11th minute. If, after this adjustment, multiple small power fluctuations (less than 0.1MW) occur around the 13th minute, the system attempts to merge these actions into a single 0.3MW charging ramp with a rate of change matching the equipment's rated ramp rate, until the final solution satisfies all constraints.

[0162] After each adjustment, the triggering reason, power value before and after the adjustment, and the change in the limit are recorded and archived. After all violation time steps are corrected, the executable power candidate scheme with constraint correction and the complete action statistics report are output as input information for cooling window control.

[0163] This invention transforms physical actions into deductible and recoverable credit resources, and intelligently correlates the credit recovery rate with real-time security status. Simultaneously, it utilizes a sliding window to accumulate short-term actions and compare them with the credit balance and window limit, achieving real-time interception of instantaneous overload risks. For any violation, the system employs an iterative correction process involving power reduction, action delay, and action merging to ensure that the output scheme maintains the scheduling objective to the greatest extent possible while strictly meeting security constraints.

[0164] S4. Implement cooling window control for the verified candidate solutions, limiting the execution time, power change amplitude, and sequence of actions, including the following steps:

[0165] S41. Based on the temperature prediction sequence in the first scheduling prediction information and combined with the rated cooling coefficient of the energy storage device, calculate the minimum cooling time required for the temperature to drop back to the safe low temperature zone after each high-power operation, and establish a corresponding cooling window within the scheduling cycle accordingly.

[0166] Example: According to the first scheduling prediction information, if a high-power discharge is predicted to occur in the 8th minute, the energy storage device temperature will rise from 35 degrees Celsius to 36 degrees Celsius. Based on the defined quantified temperature banding rules, the system determines the cooling target to reach the low-temperature band, for example, its upper limit is 34 degrees Celsius. The rated cooling coefficient specified in the equipment's technical parameters is 0.2 degrees Celsius per minute. The system calculates the minimum required cooling time, first determining that the temperature difference between the post-discharge temperature and the upper limit of the low-temperature band is 2 degrees Celsius, then dividing this temperature difference by the rated cooling coefficient to obtain a 10-minute cooling time. Therefore, the system determines that at least 10 minutes of cooling time is required after this operation to allow the predicted temperature to drop back to the safe low-temperature band range.

[0167] Furthermore, based on the minimum cooling time, continuous cooling windows are established within the scheduling cycle, specifying the start and end times of each window, and recording the minimum allowable interval and the maximum allowable power change rate.

[0168] Example: The system establishes a 10-minute cooling window for the action at the 8th minute, covering the period from the 8th to the 18th minute. During this window, the system will refer to the maximum allowable ramp rate of the equipment and consider the safety margin during the cooling period to set a more stringent power change constraint, such as setting the maximum allowable power change rate within the window to 0.3 kW per minute.

[0169] S42. After generating the cooling window, perform time screening on power change actions that have passed the action frequency constraints. If the planned execution time of an action falls within a cooling window, the system will first attempt to postpone it to the earliest executable time step after the current window ends.

[0170] If this postponement scheme results in the inability to meet the key quantitative objectives in the second scheduling information (such as instantaneous peak shaving power requirements), the system will switch to a power compromise strategy: while meeting the objectives, reduce the power of the current action to a safe value as much as possible, and recalculate its expected impact on temperature based on this reduced power value, thereby updating the current cooling window to ensure the thermal safety of the equipment.

[0171] Example: The system has established a cooling window from minute 8 to minute 18 for a high-power discharge action at minute 8. At this time, a 0.6MW discharge action is planned to be performed at minute 10 in the candidate power scheme. The system detects that the planned execution time (minute 10) is still within the limiting range of the cooling window, so it first attempts to postpone the action to the earliest executable time step after the current window ends, i.e., after minute 18 (e.g., minute 19).

[0172] If the upper-level dispatch requires a discharge power to be provided at the 11th minute to reduce peak load, the system will initiate a power compromise strategy: reducing the operating power to a safe value (e.g., 0.3MW) that meets the emergency demand while controlling temperature rise, and executing it at the 11th minute. Simultaneously, based on this new 0.3MW discharge operation, the system will re-predict the resulting temperature rise and update the current cooling window, for example, extending the original window from the 18th minute to the 20th minute, to ensure equipment thermal safety.

[0173] S43. If the power change rate of the action performed after the cooling window exceeds the rated ramp rate, it shall be divided into several continuous execution segments with single-step change rates meeting the requirements.

[0174] For actions performed after the cooling window ends, or actions performed at a new time after adjustment via S42, the system must ensure that the power change rate meets the equipment capacity. The initial planned power change rate of the current action is calculated; if it exceeds the rated ramp rate of the energy storage device, the system breaks it down into several consecutive power adjustment phases.

[0175] Further, the process is divided into several continuous execution segments with single-step change rates that meet the requirements. Each stage completes a power change within a time step that does not exceed the rated ramp rate, and there is an interval of at least one time step between stages, so that the average change rate of the entire action and the change rate of each stage are within the safe range allowed by the equipment.

[0176] Example: The system has updated its cooling window to the 20th minute. After this, the dispatch scheme requires the energy storage unit to rapidly increase its output power to respond to new load peaks. For example, the plan is to increase power from a lower level (e.g., 0.3MW) to 1.0MW at the 21st minute.

[0177] The system calculates that the power change rate for this planned action is 0.7 MW / min. Compared with the rated ramp rate specified in the equipment's technical parameters (e.g., 0.4 MW / min), the system determines that this change rate exceeds the equipment's safe operating range.

[0178] The system further activated the amplitude constraint mechanism, splitting this power increase action into two consecutive power adjustment stages: In the first stage, the power was increased from 0.3MW to 0.7MW at the 21st minute, with a change rate of 0.4MW / minute, which met the requirements; In the second stage, the power was increased from 0.7MW to the target value of 1.0MW at the 22nd minute, with a change rate of 0.3MW / minute, which also met the requirements.

[0179] When a cooling window ends, if several power change actions are triggered simultaneously at the same time, they will be executed sequentially in a fixed safety order.

[0180] Furthermore, the discharge action is performed first, followed by the charging action, and within the same type of action, the actions are performed in descending order of the change in power.

[0181] When the power change is the same, the grid-connected side power adjustment is performed first, and then the internal power adjustment of the energy storage device is performed.

[0182] In the final verification before the action is executed, the predicted values ​​of temperature and state of charge in the first scheduling prediction information are retrieved again and compared with the real-time collected temperature and state of charge measurement values. If the real-time measurement shows insufficient cooling, the current cooling window will be automatically extended. If the state of charge has not recovered to the mid-band region, the power will be reduced again until the predicted conditions are consistent with the actual measurement values.

[0183] The frequency constraints and power change decomposition within the cooling window determine the initial conditions for the sequence arrangement of multiple actions at the end of the window.

[0184] At the end of the window, the system achieves closed-loop control from prediction to final safe execution through multiple checks, including fixed-sequence execution, grid connection priority, and comparison of prediction and actual measurements. This ensures that the power changes of the energy storage device always meet safety boundaries and dispatch requirements under multiple concurrent actions.

[0185] S5. A comprehensive evaluation of the executable power candidate schemes processed by the cooling window is performed to generate the final energy storage and load coordinated scheduling command. The scheduling parameters are then adaptively adjusted based on execution feedback, including the following steps:

[0186] After the executable power candidate schemes are processed by action frequency constraints and cooling window control, the modified executable power candidate schemes are comprehensively judged and the final energy storage and load coordinated scheduling command is generated.

[0187] The comprehensive judgment includes checking whether the power amplitude, execution time and action sequence are still consistent with the future temperature, state of charge and grid-connected power prediction results of the first scheduling prediction information, and verifying whether they meet the latest second scheduling information.

[0188] During the execution of instructions, the actual charging and discharging power, state of charge, and temperature of the energy storage device are continuously collected and compared with the corresponding predicted values ​​in the first scheduling prediction information.

[0189] When any key parameter deviates from the predicted value beyond the set allowable range, the execution order is automatically corrected in the next scheduling time step, and the reason for the correction is recorded.

[0190] After completing a single correction, the final energy storage and load coordinated scheduling command is sent to the field control device, and a complete execution record is generated.

[0191] It should be further noted that key parameters include the actual charge and discharge power, state of charge, and temperature of the energy storage device.

[0192] Example 3 is an embodiment of the present invention. This embodiment provides a campus energy storage and load coordinated scheduling system with triple nested logic, including a predictive scheduling module, a candidate scheme generation module, a verification module, a cooling module, and an output module.

[0193] The predictive scheduling module acquires the operating data of the park's energy storage devices and load equipment, and generates the first scheduling prediction information;

[0194] The candidate scheme generation module generates executable power candidate schemes based on prediction information and real-time operating data, according to multi-dimensional band rules.

[0195] The verification module performs action frequency constraint processing on executable power candidate schemes, establishes an action recovery mechanism, and verifies them based on the running status.

[0196] The cooling module implements cooling window control for validated executable power candidate solutions, limiting the execution time, power change magnitude, and sequence of actions;

[0197] The output module comprehensively evaluates the executable power candidate schemes processed by the cooling window, generates the final energy storage and load coordinated scheduling instructions, and adaptively corrects the scheduling parameters based on execution feedback.

[0198] This embodiment also provides an electronic device applicable to a triple-nested logic-based campus energy storage and load collaborative scheduling method, comprising: a memory and a processor; the memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions to implement the triple-nested logic-based campus energy storage and load collaborative scheduling method proposed in the above embodiment.

[0199] This embodiment also provides a storage medium storing a computer program that, when executed by a processor, implements a triple-nested logic-based method for coordinated scheduling of energy storage and load in a park, as proposed in the above embodiment.

[0200] The storage medium proposed in this embodiment and the method for coordinated scheduling of campus energy storage and load based on a triple nested logic proposed in the above embodiments belong to the same inventive concept. Technical details not described in detail in this embodiment can be found in the above embodiments, and this embodiment has the same beneficial effects as the above embodiments.

[0201] Based on the above description of the implementation methods, those skilled in the art can clearly understand that the present invention can be implemented using software and necessary general-purpose hardware, and of course, it can also be implemented using hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as a computer floppy disk, read-only memory (ROM), random access memory (RAM), flash memory, hard disk, or optical disk, etc., including several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods of the various embodiments of the present invention.

[0202] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

Claims

1. A method for park energy storage and load collaborative scheduling of triple nested logic, characterized in that: include, Acquire operational data from energy storage devices and load equipment within the park, and generate initial scheduling prediction information; Based on predicted information and real-time operational data, executable power candidate schemes are generated according to multi-dimensional zone rules. This includes: sealing the first scheduling prediction information without direct execution, and receiving external second scheduling information; using the first scheduling prediction information as a reference, calculating the charging and discharging power range of energy storage at each time step while meeting the energy storage state of charge, temperature rise, ramp rate, and grid connection safety boundary; combining the quantitative targets for peak shaving, energy replenishment, and electricity price response in external instructions, selecting the optimal power value in each time step, and sorting the optimal planned power values ​​by time to form preliminary candidate power schemes; performing action recognition on the preliminary candidate power schemes, and constraining the preliminary candidate power schemes based on the recognition results and multi-dimensional zone rules to generate executable power candidate schemes. The multi-dimensional zone rule is a decision-making method that evaluates three types of operating parameters in real time: grid-connected power fluctuation, energy storage device temperature, and energy storage state of charge. It quantifies these parameters into discrete levels based on preset objective thresholds and dynamically calculates the maximum allowable power change of the energy storage device at the current moment based on these comprehensive levels. The system performs frequency constraint processing on executable power candidate schemes, establishes an action recovery mechanism, and verifies it based on the operating status. This includes: when the start-up, shutdown, direction switching, or power ramping action of the energy storage device is detected, deducting the corresponding credit value from the credit balance according to the power amplitude of the action and its temperature zone, state of charge zone, and grid-connected power fluctuation zone; based on the first scheduling prediction information, determining whether the temperature, state of charge, and grid-connected power fluctuation of the energy storage device meet the preset credit recovery conditions, and performing credit recovery or pausing recovery; determining a sliding window based on the cooling time in the first scheduling prediction information, accumulating statistics on the occurrence of various constraint actions within the sliding window, and comparing the statistical results with the corresponding action credit balance and window upper limit value at the current time step to determine the violation time step; performing correction operations sequentially based on the violation time step until the corresponding action credit balance and window limit are met; and outputting the corrected execution power scheme and a complete action statistics report as input information for cooling window control. Cooling window control is implemented for validated executable power candidate schemes, limiting the execution time, power change amplitude, and sequence of actions. This includes: calculating the minimum cooling time required after a high-power action based on the temperature prediction sequence in the first scheduling prediction information, combined with the rated cooling coefficient of the energy storage device and the maximum allowable ramp rate, and establishing continuous cooling windows within the scheduling cycle accordingly; if the planned execution time of the power change action in the candidate scheme falls within the cooling window, the execution time and power are adjusted to meet the cooling window constraints; if the power change rate of the action executed after the cooling window exceeds the rated ramp rate, it is divided into several continuous execution segments with single-step change rates meeting the requirements. The executable power candidate schemes processed by the cooling window are comprehensively evaluated to generate the final energy storage and load coordinated scheduling instructions, and the scheduling parameters are adaptively corrected based on the execution feedback. The comprehensive determination includes checking whether the power amplitude, execution time, and action sequence are still consistent with the future temperature, state of charge, and grid-connected power prediction results of the first scheduling prediction information, and verifying whether they meet the latest second scheduling information.

2. The park energy storage and load collaborative scheduling method of triple nested logic according to claim 1, characterized in that: The process of acquiring operational data from the park's energy storage devices and load equipment, and generating the first scheduling prediction information, includes: Voltage, current, temperature and active power acquisition units are installed at each battery cluster, grid connection interface and load branch of the energy storage device in the park. Instantaneous voltage, current, state of charge, charge and discharge rate, equipment temperature rise and load power operation data are collected through the data acquisition terminal at a set time window. The collected data is timestamped using a unified clock, and smoothing is performed on the collected data with the scheduling cycle as the time window to remove random disturbances and obtain a time-continuous running data sequence. A short-term trend prediction model is constructed based on the operational data sequence to calculate the future changes of park load power, distributed new energy output, energy storage state of charge, energy storage unit temperature and grid-connected power within the scheduling cycle; The calculated load power prediction sequence, new energy output prediction sequence, energy storage state of charge prediction sequence, energy storage temperature rise prediction sequence, and grid-connected power prediction sequence are combined to form the first scheduling prediction information, and the energy storage capacity safety boundary and power change characteristics within the prediction interval are extracted simultaneously.

3. The park energy storage and load collaborative scheduling method of triple nested logic according to claim 2, characterized in that: The step of implementing cooling window control on the verified executable power candidate schemes, limiting the execution time, power change amplitude, and sequence of actions, also includes... When a cooling window ends, if several power change actions are triggered simultaneously at the same time, they will be executed sequentially according to a fixed safety order, including: First, perform the discharging action, then the charging action, and within the same type of action, perform them in order of decreasing power change. When the power change amplitude is the same, the grid-connected side power adjustment is performed first, and then the internal power adjustment of the energy storage device is performed. In the final verification before the action is executed, the predicted values ​​of temperature and state of charge in the first scheduling prediction information are retrieved again and compared with the real-time collected temperature and state of charge measurement values. If the real-time measurement shows insufficient cooling, the current cooling window will be automatically extended. If the state of charge has not recovered to the mid-band region, the power will be reduced again until the predicted conditions are consistent with the actual measurement values.

4. The park energy storage and load collaborative scheduling method of triple nested logic according to claim 3, characterized in that: The process of comprehensively evaluating executable power candidate schemes processed by the cooling window, generating the final energy storage and load coordinated scheduling instruction, and adaptively correcting the scheduling parameters based on execution feedback includes: After the candidate power schemes have undergone action frequency constraint processing and cooling window control in sequence, the modified power schemes are comprehensively judged and the final energy storage and load coordinated scheduling command is generated. During the execution of instructions, the actual charging and discharging power, state of charge and temperature of the energy storage device are continuously collected and compared with the corresponding predicted values ​​in the first scheduling prediction information. When any key parameter deviates from the predicted value beyond the set allowable range, the execution order is automatically corrected in the next scheduling time step, and the reason for the correction is recorded. After completing a single correction, the final energy storage and load coordinated scheduling command is sent to the field control device, and a complete execution record is generated.

5. A triple nested logic park energy storage and load collaborative scheduling system, applying a triple nested logic park energy storage and load collaborative scheduling method according to any one of claims 1-4, characterized in that, include: The system includes a prediction and scheduling module, a candidate solution generation module, a verification module, a cooling module, and an output module. The predictive scheduling module acquires the operating data of the park's energy storage devices and load equipment, and generates the first scheduling prediction information; The candidate scheme generation module generates executable power candidate schemes based on prediction information and real-time operating data, in accordance with multi-dimensional band rules. The verification module performs action frequency constraint processing on executable power candidate schemes, establishes an action recovery mechanism, and verifies them based on the operating status. The cooling module implements cooling window control on the verified executable power candidate schemes, limiting the execution time, power change amplitude, and sequence of actions; The output module comprehensively evaluates the executable power candidate schemes processed by the cooling window, generates the final energy storage and load coordinated scheduling instruction, and adaptively corrects the scheduling parameters based on the execution feedback. 6.A computer device, comprising a memory and a processor, wherein the memory stores a computer program, and the computer device is configured to perform the method according to any one of claims 1-5 when the computer program is executed by the processor. When the processor executes the computer program, it implements the steps of the campus energy storage and load coordinated scheduling method according to any one of claims 1 to 4, which involves triple nested logic.

7. A computer-readable storage medium having stored thereon a computer program, characterized in that, When the computer program is executed by the processor, it implements the steps of the campus energy storage and load coordinated scheduling method of any one of claims 1 to 4, which is based on a triple nested logic.

Images