A method and system for configuring energy storage in transformer substations under multi-scenario competition
By identifying mutually exclusive time intervals, dynamically adjusting preemption weights, and implementing preemptive allocation, the problem of energy storage resource allocation deviating from actual operational needs under multi-scenario competition was solved, achieving dynamic coordination of energy storage resources and balance of energy demand.
Patent Information
- Application Number
- CN202611627524.2
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2026-10-21
- Publication Date
- 2026-11-17
AI Technical Summary
Existing technologies struggle to balance real-time operational status changes with energy demand across different scenarios under multi-scenario competition, leading to energy storage resource allocation deviating from actual operational needs.
By identifying mutually exclusive time intervals of multi-scenario competition, the preemption weight of the scenarios is dynamically adjusted, preemptive allocation is performed, and energy gap compensation is performed for the preempted scenarios, generating actual charging and discharging configuration curves under multi-scenario competition.
It enables dynamic coordination of energy storage resources in the distribution area under competitive conditions in multiple application scenarios, improves the real-time and rationality of energy storage allocation, and ensures the overall balance of energy demand in multiple scenarios in terms of time sequence.
Smart Images

Figure CN123121731A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of transformer substation energy storage operation and control technology, specifically to a transformer substation energy storage configuration method and system under multi-scenario competition. Background Technology
[0002] Energy storage in a distribution area needs to participate in multiple application scenarios simultaneously, and these scenarios differ in charging and discharging direction, power requirements, and operating periods. When multiple scenarios demand energy storage resources at the same time, conflicts may occur in power direction or capacity utilization, making it impossible for energy storage resources to meet the needs of each scenario at the same time.
[0003] Existing technologies typically employ fixed priorities or static weights for multi-scenario coordination. When the operating status of a distribution area changes, it becomes difficult to adjust the allocation strategy based on real-time safety conditions. This can easily lead to some scenarios occupying energy storage resources for extended periods, while the needs of other scenarios are continuously shelved. Furthermore, the energy demand of the preempted scenarios lacks subsequent compensation, resulting in a discrepancy between the energy storage allocation results and actual operational needs.
[0004] Therefore, existing technologies struggle to balance real-time operational status changes with energy demand across different scenarios under multi-scenario competition, and the allocation of energy storage resources is prone to deviating from the actual operational needs of the distribution area. Summary of the Invention
[0005] This invention provides a method and system for configuring energy storage in transformer substations under multi-scenario competition, which solves the problem that energy storage allocation deviates from actual operational needs when transformer substations compete for resources in multiple scenarios due to fixed allocation strategies and lack of compensation.
[0006] In a first aspect, the present invention provides a method for configuring energy storage in transformer substations under multi-scenario competition, comprising: The theoretical charging and discharging demand sequence of each application scenario of energy storage in the transformer area is obtained, and based on the power direction and target state of charge of the theoretical charging and discharging demand sequence in time sequence, the mutually exclusive time sequence intervals of multi-scenario competition are identified. Collect real-time operation data of the station area, calculate the real-time operation deviation between the real-time operation data and the preset safety boundary of each scenario, and dynamically correct the scenario preemption weight of each scenario in the mutually exclusive time interval based on the nonlinear mapping relationship of the real-time operation deviation. Within the mutually exclusive time interval, based on the corrected scenario preemption weight, the available power and available capacity of the energy storage in the transformer area are preemptively allocated to generate the actual charging and discharging configuration curve under multi-scenario competition. Extract the unmet energy gaps in the preempted scenarios, and add these unmet energy gaps to the theoretical charging and discharging demand sequence in subsequent non-mutually exclusive periods to compensate and correct the multi-scenario allocation scheme affected by competition.
[0007] Secondly, the present invention provides a transformer substation energy storage configuration system under multi-scenario competition, comprising: The mutual exclusion timing identification module is used to obtain the theoretical charging and discharging demand sequence of each application scenario of energy storage in the distribution area, and to identify the mutually exclusive timing intervals of multiple scenarios based on the power direction and target state of charge of the theoretical charging and discharging demand sequence in the timing. The preemption weight correction module is used to collect real-time operation data of the transformer area, calculate the real-time operation deviation between the real-time operation data and the preset safety boundary of each scenario, and dynamically correct the scenario preemption weight of each scenario within the mutually exclusive time interval according to the nonlinear mapping relationship of the real-time operation deviation. The preemptive allocation module is used to perform preemptive allocation of the available power and available capacity of the energy storage in the transformer area based on the corrected scenario preemption weight within the mutually exclusive time interval, and generate the actual charging and discharging configuration curve under multi-scenario competition. The energy gap compensation module is used to extract the unmet energy gaps in the preempted scenarios and add the unmet energy gaps to the theoretical charging and discharging demand sequence of subsequent non-mutually exclusive periods to compensate and correct the multi-scenario allocation scheme affected by competition.
[0008] One or more technical solutions provided in this invention have at least the following technical effects or advantages: The technical solution of this invention first obtains the theoretical charging and discharging demand sequence of peak shaving, new energy consumption, dynamic capacity expansion and emergency backup scenarios. Based on the power direction and target state of charge of each scenario in time sequence, it identifies mutual exclusion of power direction and mutual exclusion of state of charge, and obtains the mutual exclusion time sequence interval of multi-scenario competition, providing a clear conflict time sequence boundary for energy storage resource coordination.
[0009] Furthermore, real-time transformer load rate, real-time photovoltaic output value, and real-time voltage drop frequency of the power grid are collected. The real-time operational deviations of dynamic capacity expansion, new energy consumption, and emergency backup scenarios relative to the preset safety boundary are calculated respectively. The risk sensitivity coefficient is corrected by combining the exponential preemption weight function with the transformer area type label and historical mutual exclusion frequency. The preemption weight of each scenario in the mutual exclusion time interval is dynamically corrected so that the preemption priority is adaptively adjusted according to the real-time operating status of the transformer area.
[0010] Furthermore, within the mutually exclusive time interval, preemptive allocation is performed based on the corrected scenario preemption weights. When power is mutually exclusive, the scenario with the largest weight obtains the total available power of the energy storage converter. When energy is mutually exclusive, the discharge depth is controlled according to the weight dominance of peak shaving scenario and emergency backup scenario, generating the actual charge and discharge configuration curve under multi-scenario competition.
[0011] Finally, the unmet energy gaps of the preempted scenarios are extracted, labeled with scenario tags and timestamps, and stored in the competition compensation queue. In subsequent non-mutually exclusive periods, they are rolled and accumulated into the theoretical charging and discharging demand sequence according to the remaining margin and timestamp order to perform compensation charging and discharging. The gaps that have been stuck for too long are removed through the failure cycle, and the multi-scenario allocation scheme affected by competition is compensated and corrected.
[0012] In summary, the technical solution of this invention, through the synergy of four stages—mutually exclusive timing identification, dynamic preemption weight correction, preemptive allocation, and unmet energy gap compensation—enables the energy storage resources in the distribution area to be dynamically coordinated according to real-time operating status under competitive conditions in multiple application scenarios. It also compensates for affected scenarios after competition, thereby improving the real-time performance and rationality of energy storage allocation, as well as the overall temporal balance of energy demand in multiple scenarios. Attached Figure Description
[0013] Figure 1 This is a flowchart illustrating a method for configuring energy storage in a multi-scenario competitive environment, provided by the present invention.
[0014] Figure 2 This is a schematic diagram illustrating the process of generating actual charge and discharge configuration curves in a multi-scenario competitive energy storage configuration method provided by the present invention.
[0015] Figure 3 This is a schematic diagram of the structure of a transformer substation energy storage configuration method under multi-scenario competition provided by the present invention.
[0016] In the attached diagram, the labels representing each component are as follows: Mutual exclusion timing identification module 11, preemption weight correction module 12, preemptive allocation module 13, energy gap compensation module 14. Detailed Implementation
[0017] This invention provides a method and system for configuring energy storage in transformer substations under multi-scenario competition, which solves the problem that energy storage allocation deviates from actual operational needs when transformer substations compete for resources in multiple scenarios due to fixed allocation strategies and lack of compensation.
[0018] It should be noted that the terms "comprising" and "having" are intended to cover non-exclusive inclusion. For example, a process, method, system, product, or server that includes a series of steps or units is not necessarily limited to those steps or units that are explicitly listed, but may include other steps or modules that are not explicitly listed or that are inherent to these processes, methods, products, or devices.
[0019] Example 1, as Figure 1 As shown, this invention provides a method for configuring energy storage in transformer substations under multi-scenario competition, including: S100: Obtain the theoretical charging and discharging demand sequence for each application scenario of energy storage in the distribution area, and based on the power direction and target state of charge of the theoretical charging and discharging demand sequence in time sequence, identify the mutually exclusive time sequence intervals of multi-scenario competition.
[0020] This step involves obtaining the theoretical charge and discharge demand sequence for each application scenario of energy storage in the distribution area, and comparing the power direction and target state of charge of each scenario in time sequence.
[0021] In this step, the theoretical charge and discharge demand sequence for each application scenario of the energy storage area is obtained. Based on the power direction and target state of charge of the theoretical charge and discharge demand sequence in time sequence, the mutually exclusive time sequence intervals of multi-scenario competition are identified, including: The theoretical charge and discharge demand sequence for peak shaving scenarios, new energy consumption scenarios, dynamic capacity expansion scenarios, and emergency backup scenarios is loaded. When any scenario requires charging and another scenario requires discharging under the same time sequence, it is determined that power direction mutual exclusion has occurred and marked as a power mutual exclusion time sequence interval. When any scenario in the same time series requires the stored energy to be released to the first target state of charge, and another scenario requires the stored energy to be retained to the second target state of charge, which is greater than the first target state of charge, it is determined that a mutual exclusion of state of charge has occurred and is marked as an energy mutual exclusion time series interval.
[0022] Specifically, the theoretical charge and discharge demand sequences for peak shaving scenarios, new energy consumption scenarios, dynamic capacity expansion scenarios, and emergency backup scenarios are loaded first.
[0023] The theoretical charging and discharging demand sequence is generated from the load forecast of the distribution area, the output forecast of new energy sources, the time-of-use pricing period, the demand management plan, and the emergency reserve capacity configuration requirements. Each moment in the sequence includes the power demand value and the target state of charge value. The time resolution of the theoretical charging and discharging demand sequence is consistent with the control cycle of the distribution area energy storage management system, with a typical time resolution of 15 minutes, i.e., 96 points of sequence are generated daily.
[0024] The theoretical charge / discharge demand sequence is generated as follows: For peak-shaving scenarios, based on the net load curve of the distribution area and the time-of-use pricing period, the power demand value at each control point is calculated according to the principle of discharging during peak net load periods and charging during off-peak periods. For renewable energy consumption scenarios, based on the difference between the predicted distributed photovoltaic output and the distribution area load, the charging power demand value is calculated during periods when photovoltaic output exceeds load. For dynamic capacity expansion scenarios, based on the predicted distribution area load rate, the discharging power demand value is calculated during periods when the load rate exceeds the preset capacity expansion threshold. For emergency backup scenarios, constraints are generated based on the emergency backup capacity configuration requirements and a fixed target state of charge throughout the day. The above power demand values and target state of charge values are discretized according to a 15-minute control cycle to form the theoretical charge / discharge demand sequence for each scenario.
[0025] For example, taking a certain transformer substation's operating day as an example, the energy storage capacity is 500 kWh, and the rated power is 250 kW. In the peak-shaving scenario, the discharge demand is given from 19:00 to 22:00, with a power demand of 250 kW at each point and a target state of charge (SOC) of 20%. In the renewable energy consumption scenario, the charging demand is given from 12:00 to 19:00, with a power demand of 200 kW at each point. From 18:00 to 19:00, due to the reduced output of distributed photovoltaic power, the power demand is adjusted to 100 kW. In the dynamic capacity expansion scenario, the discharge demand is given from 18:00 to 21:00, with a power demand of 150 kW at each point and a target SOC of 30%. In the emergency backup scenario, the energy storage capacity is required to be no less than 40% throughout the day, i.e., the target SOC is 40%.
[0026] Secondly, when any scenario requires charging and another scenario requires discharging under the same time sequence, it is determined that power direction mutual exclusion has occurred and marked as a power mutual exclusion time sequence interval.
[0027] Among them, the determination of power direction mutual exclusion is carried out point by point according to the control cycle. At the same time point, as long as one party has a charging demand and the other party has a discharging demand, the time point is marked as a power mutual exclusion point, and consecutive power mutual exclusion points are merged into a power mutual exclusion time sequence interval.
[0028] For example, during the operating day of this transformer substation, from 18:00 to 19:00, the peak-shaving scenario has not yet entered its discharge period, while the renewable energy consumption scenario still has charging demand due to the output of distributed photovoltaic power, with a power demand of 100 kW; at the same time, the dynamic capacity expansion scenario has discharging demand due to the increased load rate of the transformer substation, with a power demand of 150 kW. At this time, the renewable energy consumption scenario requires charging, and the dynamic capacity expansion scenario requires discharging, with opposite power directions. Therefore, 18:00 to 19:00 is marked as a power mutually exclusive time interval.
[0029] Finally, when any scenario in the same time series requires the stored energy to be released to the first target state of charge, and another scenario requires the stored energy to be retained to the second target state of charge, which is greater than the first target state of charge, it is determined that a mutual exclusion of state of charge has occurred and is marked as an energy mutual exclusion time series interval.
[0030] The method for determining mutual exclusion of states of charge is as follows: at the same time point, compare the target states of charge given by each scenario. If one scenario requires the stored energy to be released to the first target state of charge, and another scenario requires the stored energy to be retained to the second target state of charge that is greater than the first target state of charge, and the scenario requiring discharge is in the discharge process, then it is determined that there is mutual exclusion of energy at that time point, and consecutive mutual exclusion points of energy are merged into mutual exclusion time intervals of energy.
[0031] For example, during a typical operating day in this transformer substation, from 19:00 to 21:00, the peak-shaving scenario requires energy storage to discharge to reduce peak load in the substation, with a target state of charge (SOC) of 20%. Simultaneously, the emergency standby scenario requires energy storage to maintain at least 40% SOC. Since the peak-shaving scenario requires energy storage to be discharged to 20%, while the emergency standby scenario requires it to remain above 40%, there is a conflict in their SOC requirements. Therefore, 19:00 to 21:00 is designated as an energy-restricted time interval. During the same period, both the peak-shaving and dynamic capacity expansion scenarios involve discharge demand, with consistent power direction, and do not constitute a power refusal. They are only designated as an energy-restricted time interval because of the conflict in their target SOCs.
[0032] The permissible state of charge (SOC) range for the energy storage system in this area is 10% to 90%. The 20% target SOC for peak shaving scenarios is 10% higher than the system's lower permissible limit, while the 40% target SOC for emergency backup scenarios is within the system's permissible operating range. Therefore, both are considered executable SOC target values for the energy storage system.
[0033] It should be noted that the above values are for illustrative purposes only and do not constitute a limitation on the present invention.
[0034] In summary, this step can identify mutually exclusive time intervals in which multiple scenarios have competing conflicts in terms of charging / discharging direction or state of charge requirements, providing clear time boundaries and conflict determination criteria for the subsequent coordinated allocation of energy storage resources.
[0035] S200: Collect real-time operation data of the station area, calculate the real-time operation deviation between the real-time operation data and the preset safety boundary of each scenario, and dynamically correct the scenario preemption weight of each scenario within the mutually exclusive time interval according to the nonlinear mapping relationship of the real-time operation deviation.
[0036] This step involves collecting real-time operational data from the transformer area and calculating the real-time operational deviation of each scenario from the preset safety boundary. It then uses an exponential nonlinear mapping to dynamically correct the preemption weight of each scenario within the mutually exclusive time interval.
[0037] In this step, real-time operational data of the transformer area is collected, and the real-time operational deviation between the real-time operational data and the preset safety boundaries for each scenario is calculated, including: Real-time load rate of transformers, real-time output value of photovoltaic power, and real-time voltage drop frequency of the power grid are collected from the measurement terminal in the distribution area. The difference between the real-time load rate of the transformer and the preset capacity expansion trigger boundary is calculated as the operating deviation of the dynamic capacity expansion scenario; The difference between the real-time photovoltaic output value and the theoretically predicted output curve is calculated as the operational deviation of the new energy consumption scenario; The difference between the real-time voltage drop frequency of the power grid and the historical normal reference frequency is calculated as the operational deviation of the emergency backup scenario.
[0038] Specifically, the real-time load rate of the transformer, the real-time output value of the photovoltaic system, and the real-time voltage drop frequency of the power grid are first collected from the measurement terminal in the distribution area.
[0039] The real-time load rate of the transformer is collected by the distribution transformer monitoring terminal in the distribution area. The collection cycle is consistent with the control cycle of the energy storage management system, typically 15 minutes. The formula for calculating the real-time load rate of the transformer is: Real-time load rate = Current active power of transformer / Rated capacity of transformer × 100%.
[0040] The real-time photovoltaic output value is transmitted by the photovoltaic grid-connected inverter or the weather power conversion module of the distribution area, and the acquisition cycle is also 15 minutes. The real-time voltage drop frequency of the power grid is counted by the power quality monitoring device of the distribution area. The statistical time window is the current control cycle, that is, the number of times the effective voltage value is lower than 90% of the rated voltage and lasts for more than 0.5 seconds within 15 minutes.
[0041] For example, a certain distribution area's energy storage management system operates on a 15-minute control cycle. In one control cycle, the distribution transformer monitoring terminal recorded a real-time transformer load rate of 85%, a real-time photovoltaic output of 120 kW from the grid-connected inverter, and the power quality monitoring device recorded one voltage drop in the current control cycle. The transformer's rated capacity is 400 kVA, and its current active power is 340 kW. Therefore, the real-time load rate is 340 kW / 400 kVA × 100% = 85%.
[0042] Secondly, the difference between the real-time transformer load rate and the preset capacity expansion trigger boundary is calculated as the operational deviation of the dynamic capacity expansion scenario. The preset capacity expansion trigger boundary is determined based on the transformer's rated capacity and the allowable overload capacity of the distribution area. When the transformer load rate has not reached the capacity expansion trigger boundary, the dynamic capacity expansion scenario has not yet been triggered, and the energy storage does not need to discharge due to the dynamic capacity expansion demand; when the transformer load rate exceeds the capacity expansion trigger boundary, the dynamic capacity expansion scenario is triggered, and the energy storage needs to discharge to reduce the transformer load rate.
[0043] A positive deviation indicates that the real-time load rate has exceeded the capacity expansion trigger boundary, and the larger the deviation, the more urgent the capacity expansion requirement. A negative deviation indicates that the real-time load rate has not yet reached the capacity expansion trigger boundary, and the dynamic capacity expansion scenario is in a non-triggered state.
[0044] For example, the preset capacity expansion trigger boundary for the transformer in this distribution area is a load rate of 80%. If the real-time load rate of the transformer in the current control cycle is 85%, then the operating deviation of the dynamic capacity expansion scenario is 85% - 80% = +5%. A positive operating deviation indicates that the transformer load rate has exceeded the capacity expansion trigger boundary, and the dynamic capacity expansion scenario is in the triggered state.
[0045] Next, the difference between the real-time photovoltaic output value and the theoretically predicted output curve is calculated as the operational deviation of the new energy consumption scenario.
[0046] The theoretical power output curve is generated by the distributed photovoltaic short-term power prediction system in the distribution area, with a prediction time resolution of 15 minutes. The real-time photovoltaic power output value is taken as the average active power transmitted by the photovoltaic grid-connected inverter within the current control cycle, which is consistent with the 15-minute time resolution of the theoretical power output curve.
[0047] A positive operational deviation indicates that the real-time output of photovoltaic power is higher than the theoretical prediction, the amount of renewable energy that can be absorbed is greater than expected, and the demand for energy storage and charging increases; a negative operational deviation indicates that the real-time output of photovoltaic power is lower than the theoretical prediction, the amount of renewable energy that can be absorbed is less than expected, and the demand for energy storage and charging decreases.
[0048] For example, if the average active power delivered by the photovoltaic grid-connected inverter during the current control cycle is 120 kW, and the predicted output value at that moment in the theoretical output curve is 100 kW, then the operational deviation of the renewable energy consumption scenario is 120 kW - 100 kW = +20 kW. A positive operational deviation indicates that the real-time photovoltaic output is higher than the theoretical prediction, and the charging demand for energy storage due to renewable energy consumption is higher than the theoretical prediction.
[0049] Finally, the difference between the real-time voltage drop frequency of the power grid and the historical normal reference frequency is calculated as the operational deviation of the emergency backup scenario.
[0050] The historical normal reference frequency is determined based on the statistical average of voltage drop frequencies in the same period over the past 30 days in the transformer area. The historical normal reference frequency is calculated as follows: sum the number of voltage drops within the same control cycle (i.e., the same 15-minute period each day) over the past 30 days, and then divide by 30 to obtain the historical normal reference frequency corresponding to that control cycle.
[0051] A positive operational deviation indicates that the frequency of grid voltage drops is higher than the historical normal level, the power quality risk in the distribution area is increased, and the demand for energy storage to ensure power supply in emergency backup scenarios is enhanced; a zero or negative operational deviation indicates that the frequency of grid voltage drops does not exceed the historical normal level, and the emergency backup scenario maintains the conventional backup constraints.
[0052] For example, if the sum of voltage dips in this distribution area within the current 15-minute period of the daily control cycle over the past 30 days is 15, then the historical normal baseline frequency is 15 times / 30 days = 0.5 times / 15 minutes. If the real-time voltage dip frequency of the power grid within the current 15-minute period is 1, then the operational deviation of the emergency backup scenario is 1 time / 15 minutes - 0.5 times / 15 minutes = +0.5 times / 15 minutes. A positive operational deviation indicates that the frequency of power grid voltage dips is higher than the historical normal level, and the power supply demand for the emergency backup scenario is enhanced.
[0053] In this step, based on the nonlinear mapping relationship of the real-time operational deviation, the scene preemption weight of each scene within the mutually exclusive time interval is dynamically adjusted, including: An exponential preemption weight function is constructed, which takes real-time running deviation as input and scene preemption weight as output. The scene preemption weight increases exponentially with the increase of real-time running deviation. Obtain the transformer area type label, which is used to identify the electricity consumption characteristic type of the transformer area, including at least one of industrial transformer areas, commercial transformer areas and residential transformer areas; The initial risk sensitivity coefficient is matched according to the type label of the transformer area. The initial risk sensitivity coefficient is used to control the response rate of the scene preemption weight as the real-time operation deviation increases. The frequency of occurrence of historical mutually exclusive time series intervals within a preset time window is counted. When the frequency of occurrence exceeds a preset fluctuation threshold, the initial risk sensitivity coefficient is amplified according to the ratio of the frequency of occurrence to the preset fluctuation threshold, and the amplified risk sensitivity coefficient does not exceed a preset upper limit value, to obtain the current risk sensitivity coefficient. The real-time operational deviation of each scenario and the current risk sensitivity coefficient are input into the exponential preemption weight function to calculate the dynamic preemption weight of each scenario in the current time series.
[0054] Specifically, an exponential preemption weight function is first constructed, which takes real-time operational deviation as input and scene preemption weight as output. The scene preemption weight increases exponentially with the increase of real-time operational deviation.
[0055] Specifically, the exponential preemption weight function takes the following form: Wi=W0×e^(λ×Δi), where Wi is the dynamic preemption weight of scenario i, W0 is the basic weight of scenario i, λ is the current risk sensitivity coefficient, and Δi is the real-time running deviation of scenario i.
[0056] The basic weight W0 is set according to the default importance of each scenario in the operation of energy storage in the distribution area, and the sum of the basic weights of each scenario is 1. The real-time operation deviation Δi is the deviation value corresponding to each scenario. The deviation of the dynamic capacity expansion scenario is the difference between the real-time load rate of the transformer and the preset capacity expansion trigger boundary. The deviation of the new energy consumption scenario is the difference between the real-time output value of photovoltaic power and the theoretical predicted output curve. The deviation of the emergency backup scenario is the difference between the real-time voltage drop frequency of the grid and the historical normal reference frequency.
[0057] The base weights are determined based on the default occupancy ratio of energy storage resources for each scenario under normal operating conditions where no real-time operational deviations trigger preemption weighting. The values are determined using either the analytic hierarchy process (AHP) or statistical analysis of historical data from transformer substations.
[0058] When using the analytic hierarchy process (AHP), a judgment matrix is constructed with the goal of improving the overall operational efficiency of energy storage in the distribution area. The relative importance of dynamic capacity expansion scenarios, new energy consumption scenarios, and emergency backup scenarios is compared pairwise. The consistency ratio of the judgment matrix should be less than 0.1. The basic weights of each scenario are obtained by solving the eigenvector method.
[0059] When using the historical data statistics method for transformer substation operation, the time integral value of the actual energy storage power or capacity occupied by each scenario in the past 30 operating days is calculated. The proportion of the time integral value of each scenario to the total time integral value of all scenarios is used as the basic weight, and normalization is performed so that the sum of the basic weights of each scenario is 1.
[0060] For example, the analytic hierarchy process (AHP) is used to determine the basic weights for this distribution area. With the overall operational efficiency of the energy storage in the distribution area as the objective, a three-scenario judgment matrix is constructed. In the judgment matrix, the importance scale of the dynamic capacity expansion scenario relative to the renewable energy consumption scenario is 2, the importance scale of the dynamic capacity expansion scenario relative to the emergency backup scenario is 3, and the importance scale of the renewable energy consumption scenario relative to the emergency backup scenario is 2. The consistency ratio of this judgment matrix is 0.017, which is less than 0.1, thus passing the consistency test. After solving for and normalizing the eigenvectors of the judgment matrix, the basic weights are obtained as follows: dynamic capacity expansion scenario: 0.539; renewable energy consumption scenario: 0.297; and emergency backup scenario: 0.164.
[0061] The rounding rule is as follows: Using a step size of 0.05, the basic weights obtained from the analytic hierarchy process are rounded to the nearest multiple of 0.05. If the sum of the three weights after rounding is not 1, the largest weight is adjusted by adding or subtracting until the sum is 1. For the dynamic capacity expansion scenario, the nearest multiple of 0.05 is 0.55; for the renewable energy consumption scenario, the nearest multiple of 0.05 is 0.30; and for the emergency backup scenario, the nearest multiple of 0.05 is 0.15. After rounding, the sum of the three weights is 1.00, resulting in a basic weight of 0.55 for the dynamic capacity expansion scenario, 0.30 for the renewable energy consumption scenario, and 0.15 for the emergency backup scenario.
[0062] Before substituting the real-time operational deviations of each scenario into the exponential preemption weight function, the real-time operational deviations of each scenario are normalized. The normalization method is as follows: divide the real-time operational deviation of each scenario by the preset normalization benchmark value corresponding to that scenario to obtain the dimensionless normalized deviation.
[0063] The preset normalization benchmark value is set as follows: for dynamic capacity expansion scenarios, the normalization benchmark value is 10% of the preset capacity expansion trigger boundary value. When the preset capacity expansion trigger boundary is 80%, the normalization benchmark value is 80% × 10% = 8%; for renewable energy consumption scenarios, the normalization benchmark value is 10% of the rated installed capacity of distributed photovoltaic power in the distribution area; for emergency backup scenarios, the normalization benchmark value is twice the historical normal benchmark frequency, and when the historical normal benchmark frequency is 0, it is taken as 0.5 times / 15 minutes. This value is the smallest non-zero statistical unit that the power quality monitoring device in the distribution area can identify, in order to avoid division by zero in the normalization calculation. The normalization deviation is directly used as the input variable of the exponential preemption weighting function.
[0064] Before normalization, the real-time operational deviations of each scenario must be uniformly converted into the same numerical expression form as the preset normalization benchmark value. The operational deviation of the dynamic capacity expansion scenario is represented by the numerical part of the load rate percentage, for example, 85%-80%=5%, and the value 5 is used for subsequent calculations; the normalization benchmark value is also represented by the numerical part, for example, 8% is represented by the value 8.
[0065] After the real-time operational deviations for each scenario are standardized in numerical representation, if the operational deviation is negative, the operational deviation for that scenario will be set to zero before normalization and substitution into the exponential preemption weight function. A zero operational deviation indicates that the scenario has not triggered a preemption request in the current control cycle, and its dynamic preemption weight is determined solely by the base weight, with the exponential function part set to 1.
[0066] For example, in this transformer substation, the base weight for the dynamic capacity expansion scenario is set to 0.55, the base weight for the renewable energy consumption scenario is set to 0.30, and the base weight for the emergency backup scenario is set to 0.15. The current risk sensitivity coefficient λ is temporarily set to 1.0; this value will only be updated and used in subsequent steps after being adjusted based on the transformer substation type and historical mutual exclusion frequency. Taking the dynamic capacity expansion scenario as an example, its current real-time operating deviation is +5%, which is taken as 5. The normalized baseline value for the dynamic capacity expansion scenario is taken as 8, and the normalized deviation is 5 ÷ 8 = 0.625. Substituting into the exponential preemption weight function, the dynamic preemption weight for the dynamic capacity expansion scenario is 0.55 × e^(1.0 × 0.625) = 0.55 × 1.868 = 1.027.
[0067] Secondly, obtain the transformer area type label, which is used to identify the electricity consumption characteristic type of the transformer area, including at least one of industrial transformer areas, commercial transformer areas, and residential transformer areas.
[0068] The historical operation records of the distribution transformer area include the daily load curves, maximum load rate, number of load rate exceedances, number of voltage dips, and distributed photovoltaic installed capacity and power generation data for the past 12 months. Based on the load characteristics and power quality characteristics in the aforementioned historical operation records, combined with the distribution transformer area classification data in the distribution network management system, the distribution transformer area type label is determined.
[0069] The historical operating records of industrial transformer substations show that the daily load curve is concentrated during production periods and the number of times the load rate exceeds the limit is relatively high; the historical operating records of commercial transformer substations show that the daily load curve is related to business hours and the load rate fluctuates greatly; the historical operating records of residential transformer substations show that the load peaks are obvious in the morning and evening and the number of times the load rate exceeds the limit is relatively low.
[0070] The different types of transformer substations mentioned above have different sensitivities to voltage quality, load rate control, and renewable energy consumption, therefore different initial risk sensitivity coefficients need to be matched.
[0071] For example, the load in this distribution area is mainly industrial equipment, and the daily load curve has obvious characteristics of concentrated production periods. The distribution area classification data in the distribution network management system identifies this distribution area as an industrial distribution area.
[0072] Next, the initial risk sensitivity coefficient is matched according to the transformer area type label. The initial risk sensitivity coefficient is used to control the response rate of the scenario preemption weight as the real-time operation deviation increases.
[0073] The initial risk sensitivity coefficient is obtained through a preset mapping table. The preset mapping table is set based on the following criteria: industrial distribution areas have a lower tolerance for voltage dips and overload rate exceedances, so the initial risk sensitivity coefficient is set higher; commercial distribution areas have a moderate sensitivity to power supply reliability, so the initial risk sensitivity coefficient is set in the middle; and residential distribution areas have a relatively higher tolerance for short-term voltage dips, so the initial risk sensitivity coefficient is set lower. The typical setting range is 0.5 to 2.0.
[0074] The initial risk sensitivity coefficient reflects the tolerance of a transformer substation for the potential negative impacts of preemptive actions. Industrial transformer substations have a low tolerance for voltage drops and overload exceeding limits caused by preemption, and therefore require a higher risk sensitivity coefficient to quickly increase the preemption weight of relevant scenarios. Residential transformer substations have a higher tolerance for the negative impacts of preemption, and their risk sensitivity coefficient can be set lower to avoid excessive fluctuations in weight.
[0075] For example, if the area is an industrial area, the initial risk sensitivity coefficient corresponding to the industrial area in the preset mapping table is 1.5.
[0076] Next, the frequency of occurrence of historical mutually exclusive time series intervals within the preset time window is counted. When the frequency of occurrence exceeds the preset fluctuation threshold, the initial risk sensitivity coefficient is amplified according to the ratio of the frequency of occurrence to the preset fluctuation threshold, and the amplified risk sensitivity coefficient does not exceed the preset upper limit value, thus obtaining the current risk sensitivity coefficient.
[0077] The preset time window is the past 7 consecutive operating days, and the statistical object is the number of control cycles that have occurred within the mutual exclusion time sequence intervals within the past 7 operating days. The statistical method for the occurrence frequency of historical mutual exclusion time sequence intervals is as follows: using the same power direction mutual exclusion determination rules and state of charge mutual exclusion determination rules as the aforementioned mutual exclusion time sequence interval identification steps, the number of power mutual exclusion control cycles and energy mutual exclusion control cycles within the past 7 operating days are combined and counted, without distinguishing between mutual exclusion types. If power mutual exclusion and energy mutual exclusion exist simultaneously within the same control cycle, it is counted only once.
[0078] A preset fluctuation threshold is used to determine the frequency of mutually exclusive time series intervals. A typical setting is 14 mutually exclusive control cycles within 7 days, averaging 2 times per day. When the frequency of historical mutually exclusive time series intervals does not exceed the preset fluctuation threshold, the current risk sensitivity coefficient is directly taken as the initial risk sensitivity coefficient without amplification. When the frequency exceeds the preset fluctuation threshold, the current risk sensitivity coefficient = initial risk sensitivity coefficient × (frequency of occurrence / preset fluctuation threshold). If the calculated result exceeds the preset upper limit, the current risk sensitivity coefficient is taken as the preset upper limit. The preset upper limit is determined based on the calculation stability of the exponential function and the response capability of the energy storage management system, typically set to 5.0.
[0079] For example, in the past 7 operating days, the number of control cycles occurring in mutually exclusive time intervals in this distribution area was 21, while the preset fluctuation threshold was 14. The frequency of occurrence exceeds the preset fluctuation threshold, with an amplification ratio of 21 / 14 = 1.5. If the initial risk sensitivity coefficient is 1.5, then the amplified risk sensitivity coefficient is 1.5 × 1.5 = 2.25. Since 2.25 does not exceed the preset upper limit of 5.0, the current risk sensitivity coefficient is determined to be 2.25.
[0080] Finally, the real-time operational deviations of each scenario and the current risk sensitivity coefficient are input into the exponential preemption weight function to calculate the dynamic preemption weight of each scenario in the current time series.
[0081] Within the mutually exclusive time interval, the dynamic preemption weight of each competing scenario is calculated cycle by cycle. The dynamic capacity expansion scenario, the renewable energy consumption scenario, and the emergency backup scenario are each substituted into the exponential preemption weight function using their respective normalized deviations and the same current risk sensitivity coefficient to obtain the dynamic preemption weight of each scenario under the current time sequence. The input variable in the exponential preemption weight function is uniformly the normalized deviation, and the product of the risk sensitivity coefficient and the normalized deviation is not subject to additional constraints.
[0082] After calculating the dynamic preemption weights for each scenario, within the same control cycle of the same mutually exclusive time interval, the dynamic preemption weights of all competing scenarios are summed. The dynamic preemption weight of each scenario is then divided by this sum to obtain the normalized dynamic preemption weight for each scenario. The sum of the normalized dynamic preemption weights is 1, which is directly used as the weight ratio of each scenario in subsequent preemptive allocation.
[0083] For example, the current risk sensitivity coefficient is 2.25. In the dynamic capacity expansion scenario, the real-time operational deviation is taken as 5, and the normalized baseline value is taken as 8. The normalized deviation is 5 / 8 = 0.625, and the dynamic preemption weight is 0.55 × e^(2.25 × 0.625) = 0.55 × 4.08 = 2.244. In the renewable energy consumption scenario, the real-time operational deviation is +20 kW. The rated installed capacity of distributed photovoltaic power in this area is 2000 kW, and the normalized baseline value is 2000 kW × 10% = 200 kW. The normalized deviation is 20 / 200 = 0.1, and the dynamic preemption weight is 0.30 × e^(2.25 × 0.1) = 0.30 × 1.25 = 0.375. The real-time operational deviation for the emergency backup scenario is +0.5 times / 15 minutes. The normalized baseline value is 0.5 times / 15 minutes × 2 = 1 time / 15 minutes, and the normalized deviation is 0.5 / 1 = 0.5. The dynamic preemption weight is 0.15 × e^(2.25 × 0.5) = 0.15 × 3.08 = 0.462. The sum of the three is 2.244 + 0.375 + 0.462 = 3.081. The normalized dynamic preemption weights are as follows: dynamic capacity expansion scenario 2.244 / 3.081 = 0.728, new energy consumption scenario 0.375 / 3.081 = 0.122, and emergency backup scenario 0.462 / 3.081 = 0.150. The sum of the normalized dynamic preemption weights is 1, which is directly used as the weight ratio of each scenario in the subsequent preemptive allocation.
[0084] It should be noted that the above values are for illustrative purposes only and do not constitute a limitation on the present invention.
[0085] In summary, this step enables the priority of energy storage resource acquisition to be adaptively adjusted according to the real-time operating status of the distribution area, thereby improving the real-time performance and rationality of energy storage allocation under multi-scenario competition.
[0086] S300: Within the mutually exclusive timing interval, based on the corrected scenario preemption weight, preemptive allocation is performed on the available power and available capacity of the energy storage in the transformer area to generate the actual charging and discharging configuration curve under multi-scenario competition.
[0087] like Figure 2 As shown, this step allocates the available energy storage power in a preemptive manner according to the modified scenario preemption weight within the mutually exclusive time interval, and controls the energy storage discharge depth according to the preemption weight dominance within the energy mutual exclusion interval, thereby generating the actual charge and discharge configuration curve under multi-scenario competition.
[0088] In this step, within the mutually exclusive timing interval, based on the corrected scenario preemption weights, a preemptive allocation is performed on the available power and available capacity of the energy storage in the distribution area, generating actual charge-discharge configuration curves under multi-scenario competition, including: When in the power mutual exclusion time interval, compare the dynamic preemption weights of each scenario, allocate the total available power of the energy storage converter in the distribution area to the scenario with the largest preemption weight, and suspend the power allocation of the other competing scenarios. When in the energy mutual exclusion time series interval, the preemption weights of peak shaving scenario and emergency backup scenario are compared. If the weight of peak shaving scenario is dominant, energy storage is allowed to discharge to the first target state of charge. If the weight of emergency backup scenario is dominant, the energy storage discharge depth is forcibly limited, and the bottom line of state of charge is raised to the second target state of charge. Based on the full allocation results and the rise in the state of charge baseline, the actual charge-discharge configuration curve is generated.
[0089] Specifically, when in the power mutual exclusion time interval, the dynamic preemption weights of each scenario are compared, and the total available power of the energy storage converter in the distribution area is fully allocated to the scenario with the largest preemption weight, while the power allocation of the other competing scenarios is suspended.
[0090] The dynamic preemption weights for each scenario are modified normalized dynamic preemption weights. Within the power mutual exclusion time interval, the normalized dynamic preemption weights of each competing scenario under the same time sequence are compared cycle by cycle, and the scenario with the largest weight is selected as the power allocation target.
[0091] The total available power of the energy storage converter in the distribution area is determined based on the current state of charge (SOC) of the energy storage system, its rated power, and the converter efficiency. When the energy storage system is in a dischargeable state, the total available power equals the smaller of the rated power of the energy storage converter and the allowable discharge power under the current SOC. When the energy storage system is in a chargeable state, the total available power equals the smaller of the rated power of the energy storage converter and the allowable charging power under the current SOC. Scenarios that are suspended do not receive a power allocation during the current control cycle, and their power demand is entered into the subsequent compensation process.
[0092] For example, during the power mutual exclusion time interval from 18:00 to 19:00, the rated power of the energy storage converter in this distribution area is 250 kW, the current state of charge is 60%, and the energy storage system is in a dischargeable state. The allowable discharge power under the current state of charge is 250 kW, therefore the total available power is 250 kW. During this control cycle, the normalized dynamic preemption weight for the renewable energy consumption scenario is 0.122, the normalized dynamic preemption weight for the dynamic capacity expansion scenario is 0.728, and the normalized dynamic preemption weight for the emergency backup scenario is 0.150. The dynamic capacity expansion scenario has the highest weight; its theoretical power demand during the current control cycle is 150 kW, and the total available power is 250 kW. Taking the smaller value of 150 kW, 150 kW is allocated to the dynamic capacity expansion scenario for discharge to reduce the transformer load rate. The renewable energy consumption scenario does not receive a charging power allocation during this control cycle.
[0093] Secondly, when in the energy mutual exclusion time series interval, the preemption weights of peak shaving scenario and emergency backup scenario are compared. If the weight of peak shaving scenario is dominant, energy storage is allowed to discharge to the first target state of charge; if the weight of emergency backup scenario is dominant, the energy storage discharge depth is forcibly restricted, and the bottom line of state of charge is raised to the second target state of charge.
[0094] The first target state of charge (SOC) is the discharge target SOC for the peak shaving scenario in the current time series, and the second target SOC is the minimum reserved SOC required for the emergency backup scenario. Within the energy mutual exclusion time series, the normalized dynamic preemption weights of the peak shaving scenario and the emergency backup scenario are compared cycle by cycle.
[0095] Specifically, when the peak-shaving scenario has a dominant weight, the lower limit of energy storage discharge is set according to the first target state of charge for the peak-shaving scenario, and the energy storage can continue to discharge until the first target state of charge. When the emergency standby scenario has a dominant weight, the lower limit of energy storage discharge is forcibly raised to the second target state of charge, and the energy storage stops discharging when it reaches the second target state of charge. When the emergency standby scenario has a dominant weight and the energy storage discharges to the second target state of charge, the energy storage stops discharging, and the discharge power is set to zero. Within the current control cycle, if there is no competing demand from other discharge scenarios in the power mutual exclusion time interval, the energy storage maintains a zero-power state; if there is competing demand from charging scenarios, charging allocation is performed according to the preemption weight of the charging scenarios.
[0096] For example, within the energy mutual exclusion time series from 19:00 to 21:00, the first target state of charge (SOC) for the peak-shaving scenario is 20%, while the second target SOC for the emergency backup scenario is 40%. Within a certain control cycle, the normalized dynamic preemption weight for the peak-shaving scenario is 0.728, and the normalized dynamic preemption weight for the emergency backup scenario is 0.150. The peak-shaving scenario has a dominant weight, so the lower limit for energy storage discharge is set at 20%, and energy storage can continue to discharge until the SOC reaches 20%.
[0097] If, after adjustment, the weight of the emergency backup scenario is 0.728 and the weight of the peak shaving scenario is 0.150 in another control cycle, the emergency backup scenario will have a dominant weight. The lower limit of energy storage discharge will be forcibly raised to 40%, and energy storage will stop discharging when it reaches 40% of its state of charge, with the discharge power set to zero. During this control cycle, there is no competing demand for charging, and energy storage will maintain a zero-power state.
[0098] Finally, based on the full allocation results and the state-of-charge (SOC) baseline elevation results, an actual charge / discharge configuration curve is generated. The actual charge / discharge configuration curve uses the control cycle as the time step, recording the actual charge / discharge power and SOC values of the energy storage system point-by-point in each control cycle. Within the non-mutually exclusive time interval, charge / discharge is executed according to the superposition of the theoretical charge / discharge demand sequences for each scenario, and the actual power value for each control cycle is recorded. Within the non-mutually exclusive time interval, if the total power demand value after the superposition of the theoretical charge / discharge demand sequences for each scenario exceeds the rated power of the energy storage converter, then the rated power of the energy storage converter is used, and the excess portion is reduced according to the basic weight ratio of each scenario. The actual allocated power for each scenario is the reduced value.
[0099] Within the power mutual exclusion time interval, the actual power value is recorded according to the preemptive allocation result. Within the energy mutual exclusion time interval, the actual power value and the state of charge limit are recorded according to the state of charge baseline control result. When the same control cycle belongs to both the power mutual exclusion time interval and the energy mutual exclusion time interval, the preemptive power allocation for the power mutual exclusion time interval is executed first to determine the power allocation object and power value for that control cycle. Then, the state of charge baseline control for the energy mutual exclusion time interval is executed to limit the discharge depth in the allocation result. If the power allocation result is a discharge scenario and the energy storage discharges to the second target state of charge, the discharge power is set to zero. The actual charging and discharging power value of each control cycle does not exceed the total available power of the energy storage converter, and the state of charge value does not exceed the allowable state of charge operating range of the energy storage system.
[0100] For example, during the operating day of this transformer substation, in the power mutual exclusion time sequence interval from 18:00 to 19:00, the actual charge / discharge configuration curve records an energy storage discharge power of 150 kW, corresponding to the discharge demand of the dynamic capacity expansion scenario. In the energy mutual exclusion time sequence interval from 19:00 to 21:00, the energy storage discharge power recorded in the control cycle with the peak-shaving scenario weight dominance is 250 kW, with a state-of-charge limit of 20%, while the energy storage discharge power recorded in the control cycle with the emergency backup scenario weight dominance is 0 kW, with a state-of-charge limit of 40%.
[0101] During non-mutually exclusive periods, such as from 12:00 to 15:00, the charging demand for new energy consumption scenarios is 200 kW. Peak shaving scenarios have no charging or discharging demand during this period. Dynamic capacity expansion scenarios are not triggered. Emergency backup scenarios only provide state of charge constraints and do not generate power demand. The total power demand after aggregation is 200 kW, which does not exceed the rated power of the energy storage converter of 250 kW. Therefore, charging is performed at 200 kW, and the energy storage charging power is recorded as 200 kW.
[0102] If a control cycle falls within both a power mutual exclusion time interval and an energy mutual exclusion time interval, and the power allocation result is peak-shaving scenario discharge while the energy mutual exclusion control result is emergency backup scenario weighted, then the discharge power of the peak-shaving scenario will be set to zero when the energy storage's state of charge reaches 40%, and discharge will cease. This ultimately generates the actual charge-discharge configuration curves for all 96 points of the operating day.
[0103] It should be noted that the above values are for illustrative purposes only and do not constitute a limitation on the present invention.
[0104] In summary, this step can generate actual charge and discharge configuration curves that take into account both real-time operating status and competition in multiple scenarios, thereby improving the real-time performance and rationality of energy storage resource allocation.
[0105] S400: Extract the unmet energy gap of the preempted scenario, and add the unmet energy gap to the theoretical charging and discharging demand sequence of subsequent non-mutually exclusive periods to compensate and correct the multi-scenario allocation scheme affected by competition.
[0106] This step extracts the unmet energy gaps generated in the preempted scenario within the mutually exclusive time interval, and adds them to the theoretical charging and discharging demand sequence of subsequent non-mutually exclusive time periods in order of timestamps to perform compensation. At the same time, a failure cycle is set to remove gaps that have been stuck for too long.
[0107] In this step, the unmet energy gaps in the preempted scene are extracted, including: The difference between the theoretical charge or discharge energy and the actual energy executed in scenarios where power allocation is deprived or discharge depth is limited within the mutually exclusive timing interval is calculated as the unmet energy gap. The unmet energy gaps are marked with scene tags and timestamps and stored in the competition compensation queue.
[0108] Specifically, the difference between the theoretical charge or discharge energy and the actual execution energy is first calculated for scenarios where the power allocation is deprived or the discharge depth is limited within the mutually exclusive timing interval, and this difference is taken as the unmet energy gap.
[0109] For scenarios where power allocation is deprived, the theoretical charge or discharge energy is equal to the product of the theoretical power demand in the current control cycle and the control cycle duration. The actual execution energy is equal to the product of the actual power obtained in the current control cycle and the control cycle duration. When the actual power obtained is zero, the actual execution energy is zero.
[0110] For scenarios with limited discharge depth, the theoretically expected energy release equals the product of the theoretical discharge power value and the control cycle duration within the current control cycle. The actual energy released equals the product of the actual discharge power value and the actual discharge duration within the current control cycle. The actual discharge duration is calculated from the start of the current control cycle until the energy storage state of charge reaches the second target state of charge, and this duration does not exceed the control cycle duration. The actual energy released equals the energy actually released after prematurely stopping discharge due to the rise in the state of charge threshold. The control cycle duration is the control cycle of the energy storage management system, typically 15 minutes, or 0.25 hours.
[0111] For example, during the power mutual exclusion time interval from 18:00 to 19:00, the power allocation quota for the renewable energy consumption scenario is revoked. In this scenario, the theoretical charging power demand within the current control cycle is 100 kW, the control cycle duration is 0.25 hours, and the theoretical charging energy should be 100 kW × 0.25 hours = 25 kWh. The actual power received is 0 kW, and the actual executed energy is 0 kWh. The unmet energy gap is 25 kWh - 0 kWh = 25 kWh, representing a charging gap.
[0112] Furthermore, unmet energy gaps are marked with scene tags and timestamps and stored in the competition compensation queue.
[0113] Scenario tags identify the application scenario to which the unmet energy gap belongs. Timestamps record the start time of the control cycle in which the unmet energy gap occurs, expressed as the start time of the control cycle, accurate to the minute, in the format of year-month-day-hour:minute. The competition compensation queue stores each unmet energy gap in chronological order of its timestamp. Within the same control cycle, unmet energy gaps from multiple scenarios are arranged in descending order of their basic scenario weight. Each record in the queue includes a scenario tag, a timestamp, the value of the unmet energy gap, and the gap direction (either a charging gap or a discharging gap).
[0114] For example, the unmet energy gap in the aforementioned renewable energy consumption scenario is tagged as a renewable energy consumption scenario, with a timestamp of 18:00, an unmet energy gap value of 25 kWh, and a gap direction of charging gap. This record is stored in the competition compensation queue. During the energy mutual exclusion time series from 19:00 to 21:00, if a peak-shaving scenario is restricted in its discharge depth due to the priority of the emergency backup scenario, its unmet energy gap is tagged as a peak-shaving scenario in the same way, with a timestamp of the corresponding control cycle start time, a gap direction of discharging gap, and stored in the competition compensation queue in timestamp order.
[0115] In a peak-shaving scenario, during a control cycle from 19:00 to 21:00, the theoretical discharge power is 250 kW, the control cycle duration is 0.25 hours, and the theoretical energy to be discharged is 62.5 kWh. Because the emergency backup scenario has a higher weighting, the energy storage stops discharging when it reaches 40% state of charge. The actual discharge duration is 0.1 hours, the actual discharge power is 250 kW, and the actual energy executed is 250 kW × 0.1 hours = 25 kWh. The unmet energy gap is 62.5 kWh - 25 kWh = 37.5 kWh, which is the discharge gap.
[0116] In this step, the unmet energy gap is rolled over and accumulated into the theoretical charge / discharge demand sequence of subsequent non-mutually exclusive periods to compensate and correct the multi-scenario allocation scheme affected by competition, including: In subsequent non-mutually exclusive periods, when there is a margin of available energy storage power, unmet energy gaps are extracted from the competition compensation queue in chronological order of timestamps and added to the theoretical charging and discharging demand sequence of the current period to perform compensation charging and discharging. When the compensated charging and discharging conflict with the theoretical charging and discharging demand for the current period, the demand for the current period takes priority, and the uncompensated gap remains in the competitive compensation queue to wait for the next compensation opportunity. The failure period is matched according to the scenario label of the unmet energy gap, wherein the first failure period corresponding to the emergency backup scenario is shorter than the second failure period corresponding to the peak shaving scenario. When the unmet energy gap remains in the competitive compensation queue for longer than the corresponding failure period, it is removed from the queue, and compensation correction is stopped.
[0117] Specifically, in subsequent non-mutually exclusive periods, when there is a margin of available energy storage power, unmet energy gaps are extracted from the competitive compensation queue in chronological order of timestamps and added to the theoretical charging and discharging demand sequence of the current period to perform compensation charging and discharging.
[0118] The remaining margin of available energy storage power equals the difference between the total available power of the energy storage converter and the total power demand value after superimposing the theoretical charging and discharging demand sequence for the current period. When the remaining margin is greater than zero, it indicates that the energy storage system has additional power space available for compensation within the current control cycle. When extracting unmet energy gaps, they are extracted sequentially from the head of the queue according to their timestamps. The compensation power value corresponding to an unmet energy gap extracted in a single instance does not exceed the remaining margin of the current control cycle. The power direction of compensation charging and discharging is consistent with the gap direction of the unmet energy gap; charging gaps are compensated by charging, and discharging gaps are compensated by discharging. After compensation is executed, the value of the unmet energy gap is reduced accordingly, and the compensated portion is subtracted from the queue record.
[0119] After compensation charging and discharging is performed, the energy storage state of charge (SBC) is updated according to the actual charging or discharging amount. During charging compensation, the increase in SBC is the actual charging energy divided by the rated energy storage capacity; during discharging compensation, the decrease in SBC is the actual discharging energy divided by the rated energy storage capacity. The updated SBC serves as the input for calculating the total available power of energy storage in the next control cycle. Within the same control cycle, extraction stops when the remaining margin is exhausted or the queue is empty.
[0120] For example, during the non-mutually exclusive period after 19:00, the total available power of the energy storage converter within a certain control cycle is 250 kW. The total power demand after superimposing the theoretical charging and discharging demand sequence for the current period is 150 kW, leaving a margin of 250 kW - 150 kW = 100 kW. The earliest timestamp record in the competitive compensation queue is the charging gap for the new energy consumption scenario at 18:00, with an unmet energy gap value of 25 kWh, corresponding to a compensation power of 25 kWh / 0.25 hours = 100 kW. Adding the 100 kW charging compensation to the theoretical charging and discharging demand sequence for the current period, the actual executed charging power is 150 kW + 100 kW = 250 kW. After the compensation is executed, the unmet energy gap value is reduced to 0 and removed from the queue.
[0121] Furthermore, when the compensated charging / discharging conflict with the theoretical charging / discharging demand for the current period, the demand for the current period takes priority, and the uncompensated gap remains in the competitive compensation queue waiting for the next compensation opportunity.
[0122] The judgment condition for the conflict between the compensation charge and discharge and the theoretical charge and discharge demand of the current period is: within the same control cycle, the power direction of the compensation charge and discharge is opposite to the power direction of any scenario in the theoretical charge and discharge demand sequence of the current period, or the compensation charge and discharge causes the energy storage state of charge to exceed the allowable operating range.
[0123] In the aforementioned conflict scenario, the theoretical charging and discharging demand for the current period will be prioritized, while compensation charging and discharging will not be implemented for the time being. Uncompensated gaps will not be deducted but will remain in the competitive compensation queue, awaiting reassessment of compensation conditions in subsequent control cycles.
[0124] For example, in a subsequent non-mutually exclusive time period, the theoretical charging and discharging demand for the current period is 150 kW of discharge for peak-shaving scenarios, while the compensation gap in the competitive compensation queue is the charging gap for renewable energy consumption scenarios, with the compensation direction being charging. The charging compensation and the current period's discharge demand power direction are opposite, creating a conflict. In this case, the 150 kW discharge for peak-shaving scenarios in the current period is executed first, while the charging compensation gap for renewable energy consumption scenarios remains in the queue, waiting for subsequent periods with charging demand or periods with remaining margin allowing for charging before compensation is executed.
[0125] Furthermore, the corresponding failure period is matched according to the scenario label of the unmet energy gap, wherein the first failure period corresponding to the emergency backup scenario is shorter than the second failure period corresponding to the peak shaving scenario.
[0126] The failure period is used to limit the maximum time that unmet energy gaps remain in the competitive compensation queue. The failure period is set based on the following: the power supply requirements in emergency backup scenarios are time-sensitive, and the significance of compensation decreases after the risk of voltage drop is eliminated, so a shorter first failure period is set; the peak load periods in peak shaving scenarios are predictable and substitutable, and the compensation window can be appropriately widened, so a longer second failure period is set.
[0127] The typical first failure cycle is 2 hours, and the typical second failure cycle is 4 hours. The failure cycles for renewable energy consumption and dynamic capacity expansion scenarios can be set according to the actual operational needs of the transformer substation. The failure cycle setting for renewable energy consumption scenarios is based on the fact that photovoltaic output prediction errors increase with the prediction time scale, and the reliability of compensation charging decreases significantly after 3 hours; therefore, a typical value of 3 hours is used. The failure cycle setting for dynamic capacity expansion scenarios is based on the fact that the duration of transformer load factor exceeding limits usually does not exceed 2 hours. After 2 hours, the capacity expansion demand has been eliminated, and compensation discharge is no longer necessary; therefore, a typical value of 2 hours is used.
[0128] For example, the first failure period for the emergency backup scenario in this area is set to 2 hours, and the second failure period for the peak shaving scenario is set to 4 hours. In the competitive compensation queue, the charging gap in the new energy consumption scenario generated at 18:00 is calculated according to the failure period of 3 hours for the new energy consumption scenario, with the failure time being 21:00. The discharge gap in the peak shaving scenario generated between 19:00 and 21:00 is calculated according to the failure period of 4 hours for the peak shaving scenario, with the failure time being 4 hours after the corresponding timestamp.
[0129] Finally, when the unmet energy gap remains in the competitive compensation queue for longer than the corresponding failure period, it is removed from the queue, and compensation correction is stopped.
[0130] The dwell time is calculated from the timestamp of the energy gap not being met until the start of the current control cycle. The dwell time of each record in the queue is checked cycle by cycle. When the dwell time exceeds the failure period of the corresponding scenario tag, the record is removed from the queue and no further compensation charging / discharging is performed. After removal, the scenario corresponding to the unmet energy gap will no longer receive subsequent compensation.
[0131] For example, a charging gap in a renewable energy consumption scenario that occurs at 18:00 has a failure period of 3 hours, with the failure time being 21:00. If the gap is not compensated by 21:00, and the delay time reaches 3 hours, exceeding the failure period, then the record will be removed from the competitive compensation queue, and compensation and correction for the energy preempted in that renewable energy consumption scenario will cease.
[0132] It should be noted that the above values are for illustrative purposes only and do not constitute a limitation on the present invention.
[0133] In summary, this step can reduce the impact of multi-scenario competition on the energy demand of the preempted scenario and improve the overall temporal balance of the energy storage resource allocation scheme.
[0134] In this invention, after generating the actual charge / discharge configuration curves under multi-scenario competition, the method further includes: Obtain the theoretical benefit index and theoretical safety index of the preempted scenario within the mutually exclusive time interval. The theoretical benefit index is the expected benefit value calculated based on the theoretical charging and discharging requirements of each scenario. The theoretical safety index is the expected safety margin value calculated based on the preset safety boundary of each scenario. Calculate the actual benefit index and actual safety index after the preemption. The actual benefit index is the actual benefit value calculated based on the actual charge and discharge configuration curve, and the actual safety index is the actual safety margin value calculated based on the actual charge and discharge configuration curve. The difference between the theoretical and actual return indicators is taken as the negative impact of economic loss, and the difference between the theoretical and actual safety indicators is taken as the negative impact of deterioration of safety margin. Based on the type label of the transformer area, the corresponding negative impact tolerance threshold is retrieved. When the negative impact of economic loss or the negative impact of deterioration of security margin exceeds the negative impact tolerance threshold, an overzealous preemption strategy alarm is generated.
[0135] Specifically, the theoretical revenue and theoretical safety metrics for the preempted scenario within the mutually exclusive time interval are first obtained. The theoretical revenue metric is the expected revenue value calculated based on the theoretical charging and discharging requirements of each scenario, and the theoretical safety metric is the expected safety margin value calculated based on the preset safety boundaries of each scenario.
[0136] The theoretical benefit indicators are calculated separately for each scenario. For peak-shaving scenarios, the theoretical benefit indicator is calculated by multiplying the theoretical discharge capacity by the peak-valley electricity price, i.e., the theoretical discharge capacity multiplied by the difference between the peak-hour electricity price and the valley-hour charging cost. For renewable energy consumption scenarios, the theoretical benefit indicator is calculated by multiplying the theoretical charging capacity by the renewable energy consumption subsidy or curtailment penalty. For dynamic capacity expansion scenarios, the theoretical benefit indicator is calculated based on the reduced demand electricity cost or transformer expansion cost due to the theoretical discharge capacity. For emergency backup scenarios, the theoretical benefit indicator is calculated based on the reliability benefit converted from the theoretical power reserve capacity.
[0137] Specifically, theoretical safety indicators are calculated separately for each scenario. For dynamic capacity expansion scenarios, the theoretical safety indicator is the difference between the preset capacity expansion trigger boundary and the theoretical load rate, i.e., the theoretical safety margin value. For emergency backup scenarios, the theoretical safety indicator is the difference between the theoretical state of charge and the minimum reserved state of charge. For renewable energy consumption scenarios, the theoretical safety indicator is the difference between the theoretical consumption rate and the preset lower limit of the consumption rate. For peak shaving scenarios, the theoretical safety indicator is the difference between the theoretical peak shaving completion rate and the preset lower limit of the peak shaving completion rate.
[0138] For example, in the power mutual exclusion time sequence interval from 18:00 to 19:00, the dynamic capacity expansion scenario in this distribution area is allocated a discharge power of 150 kW, with a control cycle duration of 0.25 hours and a theoretical discharge amount of 37.5 kWh. In the peak shaving scenario, the discharge depth is limited in the energy mutual exclusion time sequence interval from 19:00 to 21:00, and the theoretical discharge amount in a certain control cycle is 62.5 kWh. Taking the dynamic capacity expansion scenario as an example, the theoretical revenue indicator is calculated based on the reduction in demand electricity price. Assuming the demand electricity price is 40 yuan / kW·month, the theoretical revenue per discharge is 37.5 kWh × 0.5 yuan / kWh = 18.75 yuan. The theoretical safety indicator for the dynamic capacity expansion scenario is the difference between the preset capacity expansion trigger boundary of 80% and the theoretical load rate of 82%, i.e., -2%. This indicates that under theoretical conditions, the load rate has exceeded the capacity expansion trigger boundary, requiring energy storage discharge to reduce the load rate below 80%.
[0139] Secondly, calculate the actual benefit index and actual safety index after the preemption. The actual benefit index is the actual benefit value calculated based on the actual charge-discharge configuration curve, and the actual safety index is the actual safety margin value calculated based on the actual charge-discharge configuration curve.
[0140] The calculation method for actual revenue indicators is the same as that for theoretical revenue indicators, but the charging and discharging volume is taken as the actual electricity value executed in each scenario in the actual charging and discharging configuration curve. The calculation method for actual safety indicators is the same as that for theoretical safety indicators, but parameters such as load rate, state of charge, absorption rate, and peak shaving completion rate are taken as the actual values after the actual charging and discharging configuration curve is executed.
[0141] For example, in the dynamic capacity expansion scenario, the actual discharge power obtained between 18:00 and 19:00 was 150 kW, with an actual discharge of 37.5 kWh, consistent with the theoretical discharge, resulting in an actual revenue of 18.75 yuan. The actual safety indicator, the difference between the preset capacity expansion trigger boundary of 80% and the actual load rate after energy storage discharge of 82%, is still -2%, indicating that the load rate after energy storage discharge still exceeds the capacity expansion trigger boundary. In the peak shaving scenario, the discharge depth was limited during a certain control cycle between 19:00 and 21:00 due to the priority of the emergency backup scenario, resulting in an actual discharge of 0 kWh and an actual revenue of 0 yuan.
[0142] Furthermore, the difference between theoretical and actual return indicators is considered as the negative impact of economic loss, and the difference between theoretical and actual safety indicators is considered as the negative impact of deterioration in safety margin.
[0143] The negative impact of economic loss is calculated as the positive value obtained by subtracting the actual benefit from the theoretical benefit, representing the economic benefits lost due to the preemption of the scenario. When the actual benefit is greater than or equal to the theoretical benefit, the negative impact of economic loss is zero. The negative impact of security margin deterioration is calculated as the positive value obtained by subtracting the actual security indicator from the theoretical security indicator, representing the extent to which the security margin decreases due to the preemption of the scenario. When the actual security indicator is greater than or equal to the theoretical security indicator, the negative impact of security margin deterioration is zero.
[0144] For example, in a peak-shaving scenario, the theoretical revenue for a certain control cycle from 19:00 to 21:00 is 62.5 kWh × 0.7 yuan / kWh = 43.75 yuan, while the actual revenue is 0 yuan, resulting in a negative economic loss of 43.75 yuan. In a dynamic capacity expansion scenario, the theoretical safety margin is -2%, the actual safety margin is -2%, and the negative impact of deteriorated safety margin is 0%.
[0145] Finally, the corresponding negative impact tolerance threshold is retrieved based on the type label of the transformer area. When the negative impact of economic loss or the negative impact of deterioration of security margin exceeds the negative impact tolerance threshold, an overzealous preemption strategy alarm is generated.
[0146] After generating an overreaction alarm for the preemption strategy, the local control terminal of the energy storage area will temporarily reduce the current risk sensitivity coefficient to 50% of the initial risk sensitivity coefficient in the next control cycle to suppress overreactions during subsequent mutually exclusive time intervals. The temporary reduction will continue until the end of the day's operation, and the corrected initial risk sensitivity coefficient will be used again the following day.
[0147] The negative impact tolerance threshold is obtained through a preset mapping table, which sets the economic loss tolerance threshold and the safety margin deterioration tolerance threshold according to the transformer substation type label. Industrial transformer substations have a lower tolerance for safety margin deterioration, so the safety margin deterioration tolerance threshold is set smaller; residential transformer substations have a relatively higher tolerance for economic losses, so the economic loss tolerance threshold is set larger.
[0148] For example, in a typical setup, the economic loss tolerance threshold for industrial transformer substations is 50 yuan per control cycle, and the safety margin deterioration tolerance threshold is 3%; the economic loss tolerance threshold for commercial transformer substations is 40 yuan per control cycle, and the safety margin deterioration tolerance threshold is 2%; and the economic loss tolerance threshold for residential transformer substations is 30 yuan per control cycle, and the safety margin deterioration tolerance threshold is 2%.
[0149] For example, this distribution area is an industrial distribution area, with an economic loss tolerance threshold of 50 yuan per control cycle and a safety margin deterioration tolerance threshold of 3%. The negative economic loss in the peak-shaving scenario is 43.75 yuan, which does not exceed 50 yuan, so no excessive economic loss alarm is generated. The negative impact of safety margin deterioration in the dynamic capacity expansion scenario is 0%, which does not exceed 3%, so no excessive safety margin deterioration alarm is generated.
[0150] If the negative economic loss reaches 60 yuan within a certain control cycle, exceeding the 50 yuan threshold, an overzealous preemption strategy alarm will be generated, indicating that the current preemption strategy has an excessive impact on the economic loss of the preempted scenario. The local control terminal of the energy storage area will temporarily reduce the current risk sensitivity coefficient from 2.25 to 1.125 in the next control cycle, and this will continue until the end of the day's operation.
[0151] This invention also includes: The frequency of daily mutual exclusion time intervals, the preemption ratio of each scenario, the negative impact of economic losses, and the negative impact of deteriorated security margins are packaged and uploaded to the cloud platform. The ratio of the difference between the actual operating deviation of each scenario after compensation and correction and the theoretical operating deviation before compensation and correction to the theoretical operating deviation is calculated as the improvement rate. When the improvement rate is lower than the preset improvement lower limit, it is determined that the current risk sensitivity coefficient is too low, and a risk sensitivity coefficient correction command is generated and sent to the local control terminal of the energy storage in the distribution area to increase the initial risk sensitivity coefficient.
[0152] Specifically, the frequency of daily mutual exclusion time intervals, the preemption ratio of each scenario, the negative impact of economic losses, and the negative impact of deteriorated security margins are packaged and uploaded to the cloud platform.
[0153] The frequency of mutual exclusion time series intervals is the sum of the number of power mutual exclusion control cycles and the number of energy mutual exclusion control cycles identified within the past operating day. The occurrence of two types of mutual exclusion within the same control cycle is counted as one instance. The preemption ratio for each scenario is the proportion of the number of control cycles in which each scenario actually receives power allocation within the mutual exclusion time series interval to the total number of control cycles in the mutual exclusion time series interval. The negative impact of economic loss is the cumulative value of the negative impact of economic loss for each control cycle on that day, and the negative impact of deterioration of safety margin is the maximum value of the negative impact of deterioration of safety margin for each control cycle on that day. Uploaded data is packaged by operating day, including the transformer area identifier, date, and values of each indicator, and uploaded to the cloud platform through the communication module of the local control terminal of the transformer area's energy storage.
[0154] For example, in one operating day, the frequency of mutually exclusive timing intervals in this distribution area is 5 times, including 3 power mutual exclusions and 2 energy mutual exclusions. The number of control cycles that actually obtain power allocation in each scenario within the mutually exclusive timing intervals are as follows: dynamic capacity expansion scenario 3 times, peak shaving scenario 1 time, renewable energy consumption scenario 0 times, and emergency backup scenario 1 time, with corresponding preemption ratios of 60%, 20%, 0%, and 20%, respectively. The cumulative negative impact of economic loss on that day is 120 yuan, and the maximum negative impact of deterioration in safety margin is 1.5%. The above data is packaged and uploaded to the cloud platform.
[0155] Furthermore, the ratio of the difference between the actual operating deviation after compensation and the theoretical operating deviation before compensation and correction to the theoretical operating deviation is calculated as the improvement rate. When the improvement rate is lower than the preset lower limit of improvement, the current risk sensitivity coefficient is determined to be too low.
[0156] The theoretical operating deviation before compensation correction is the real-time operating deviation of each scenario within the mutually exclusive time interval. The actual operating deviation after compensation correction is the average value of the operating deviations recalculated for each scenario between the end of the mutually exclusive time interval that generates the unmet energy gap and the end of the day's operation after the compensation charging and discharging is completed.
[0157] The improvement rate is calculated as follows: Improvement Rate = (Theoretical Operating Deviation Before Compensation Correction - Actual Operating Deviation After Compensation Correction) / Theoretical Operating Deviation Before Compensation Correction × 100%. When the theoretical operating deviation before compensation correction is 0, this scenario is not included in the improvement rate calculation. The preset improvement lower limit is set according to the transformer area type label, with a typical value of 10%.
[0158] For example, in a dynamic capacity expansion scenario, an unmet energy gap occurs during the mutually exclusive time interval from 18:00 to 19:00. Compensation charging and discharging are performed during the non-mutually exclusive period after 19:00. The actual operating deviation after compensation correction is the average of the operating deviations of each control cycle in the dynamic capacity expansion scenario from 19:00 to the end of the day's operation, calculated as +2%. The improvement rate is (5%-2%) / 5%×100%=60%. 60% is higher than the preset improvement lower limit of 10%, indicating that the current risk sensitivity coefficient is not too low. If the improvement rate of the new energy consumption scenario is 5%, which is lower than the preset improvement lower limit of 10%, then the current risk sensitivity coefficient is considered too low.
[0159] Finally, when the improvement rate is lower than the preset improvement limit, a risk sensitivity coefficient correction instruction is generated and sent to the local control terminal of the energy storage area to increase the initial risk sensitivity coefficient.
[0160] The risk sensitivity coefficient correction instruction includes the correction target, correction direction, and correction step size. The correction target is the initial risk sensitivity coefficient, and the correction direction is to increase it. The correction step size is determined based on the difference between the improvement rate and the preset improvement lower limit; the larger the difference, the larger the correction step size, with a typical correction step size of 10% to 30% of the initial risk sensitivity coefficient.
[0161] After receiving the correction command, the local control terminal of the energy storage area increases the initial risk sensitivity coefficient by the correction step size and updates it in the preemption weight calculation for subsequent operating days. The increased initial risk sensitivity coefficient shall not exceed the preset upper limit of 5.0. If the increased initial risk sensitivity coefficient exceeds the preset upper limit, the preset upper limit shall be used.
[0162] For example, the initial risk sensitivity coefficient is currently 1.5. The improvement rate for the renewable energy consumption scenario is 5%, with a preset lower limit of improvement of 10%, resulting in a difference of 5 percentage points. The correction step size is set to 20% of the initial risk sensitivity coefficient, i.e., 1.5 × 20% = 0.3. The corrected initial risk sensitivity coefficient is 1.5 + 0.3 = 1.8. Since 1.8 does not exceed the preset upper limit of 5.0, a correction command is issued to the local control terminal of the energy storage area, updating the initial risk sensitivity coefficient to 1.8. Subsequent operating days will use 1.8 as the initial risk sensitivity coefficient for preemptive weight calculations.
[0163] It should be noted that the above values are for illustrative purposes only and do not constitute a limitation on the present invention.
[0164] In summary, this invention, through the synergy of four stages—mutually exclusive timing identification, dynamic preemption weight correction, preemptive allocation, and unmet energy gap compensation—enables the energy storage resources in the distribution area to be dynamically coordinated according to real-time operating status under competitive conditions in multiple application scenarios. It also compensates for affected scenarios after competition, thereby improving the real-time performance and rationality of energy storage allocation, as well as the overall temporal balance of energy demand in multiple scenarios.
[0165] Example 2, as Figure 3 As shown, this invention provides a transformer substation energy storage configuration system under multi-scenario competition, comprising: The mutual exclusion timing identification module 11 is used to obtain the theoretical charging and discharging demand sequence of each application scenario of energy storage in the distribution area, and to identify the mutually exclusive timing intervals of multiple scenarios based on the power direction and target state of charge of the theoretical charging and discharging demand sequence in the timing.
[0166] This includes obtaining the theoretical charge and discharge demand sequence for each application scenario of energy storage in the distribution area, and based on the power direction and target state of charge of the theoretical charge and discharge demand sequence in time sequence, identifying mutually exclusive time sequence intervals for multi-scenario competition, including: The theoretical charge and discharge demand sequence for peak shaving scenarios, new energy consumption scenarios, dynamic capacity expansion scenarios, and emergency backup scenarios is loaded. When any scenario requires charging and another scenario requires discharging under the same time sequence, it is determined that power direction mutual exclusion has occurred and marked as a power mutual exclusion time sequence interval. When any scenario in the same time series requires the stored energy to be released to the first target state of charge, and another scenario requires the stored energy to be retained to the second target state of charge, which is greater than the first target state of charge, it is determined that a mutual exclusion of state of charge has occurred and is marked as an energy mutual exclusion time series interval.
[0167] The preemption weight correction module 12 is used to collect real-time operation data of the station area, calculate the real-time operation deviation between the real-time operation data and the preset safety boundary of each scenario, and dynamically correct the scenario preemption weight of each scenario within the mutually exclusive time interval according to the nonlinear mapping relationship of the real-time operation deviation.
[0168] This includes collecting real-time operational data from the data acquisition area and calculating the real-time operational deviation between the real-time operational data and the preset safety boundaries for each scenario, including: Real-time load rate of transformers, real-time output value of photovoltaic power, and real-time voltage drop frequency of the power grid are collected from the measurement terminal in the distribution area. The difference between the real-time load rate of the transformer and the preset capacity expansion trigger boundary is calculated as the operating deviation of the dynamic capacity expansion scenario; The difference between the real-time photovoltaic output value and the theoretically predicted output curve is calculated as the operational deviation of the new energy consumption scenario; The difference between the real-time voltage drop frequency of the power grid and the historical normal reference frequency is calculated as the operational deviation of the emergency backup scenario.
[0169] Based on the nonlinear mapping relationship of the real-time operational deviation, the scene preemption weight of each scene within the mutually exclusive time interval is dynamically adjusted, including: An exponential preemption weight function is constructed, which takes real-time running deviation as input and scene preemption weight as output. The scene preemption weight increases exponentially with the increase of real-time running deviation. Obtain the transformer area type label, which is used to identify the electricity consumption characteristic type of the transformer area, including at least one of industrial transformer areas, commercial transformer areas and residential transformer areas; The initial risk sensitivity coefficient is matched according to the type label of the transformer area. The initial risk sensitivity coefficient is used to control the response rate of the scene preemption weight as the real-time operation deviation increases. The frequency of occurrence of historical mutually exclusive time series intervals within a preset time window is counted. When the frequency of occurrence exceeds a preset fluctuation threshold, the initial risk sensitivity coefficient is amplified according to the ratio of the frequency of occurrence to the preset fluctuation threshold, and the amplified risk sensitivity coefficient does not exceed a preset upper limit value, to obtain the current risk sensitivity coefficient. The real-time operational deviation of each scenario and the current risk sensitivity coefficient are input into the exponential preemption weight function to calculate the dynamic preemption weight of each scenario in the current time series.
[0170] The preemptive allocation module 13 is used to perform preemptive allocation of the available power and available capacity of the energy storage in the transformer area based on the corrected scenario preemption weight within the mutually exclusive time interval, and generate the actual charge and discharge configuration curve under multi-scenario competition.
[0171] Within the mutually exclusive time interval, based on the corrected scenario preemption weights, a preemptive allocation is performed on the available power and available capacity of the energy storage in the distribution area, generating actual charge-discharge configuration curves under multi-scenario competition, including: When in the power mutual exclusion time interval, compare the dynamic preemption weights of each scenario, allocate the total available power of the energy storage converter in the distribution area to the scenario with the largest preemption weight, and suspend the power allocation of the other competing scenarios. When in the energy mutual exclusion time series interval, the preemption weights of peak shaving scenario and emergency backup scenario are compared. If the weight of peak shaving scenario is dominant, energy storage is allowed to discharge to the first target state of charge. If the weight of emergency backup scenario is dominant, the energy storage discharge depth is forcibly limited, and the bottom line of state of charge is raised to the second target state of charge. Based on the full allocation results and the rise in the state of charge baseline, the actual charge-discharge configuration curve is generated.
[0172] After generating actual charge / discharge configuration curves under multi-scenario competition, the following is also included: Obtain the theoretical benefit index and theoretical safety index of the preempted scenario within the mutually exclusive time interval. The theoretical benefit index is the expected benefit value calculated based on the theoretical charging and discharging requirements of each scenario. The theoretical safety index is the expected safety margin value calculated based on the preset safety boundary of each scenario. Calculate the actual benefit index and actual safety index after the preemption. The actual benefit index is the actual benefit value calculated based on the actual charge and discharge configuration curve, and the actual safety index is the actual safety margin value calculated based on the actual charge and discharge configuration curve. The difference between the theoretical and actual return indicators is taken as the negative impact of economic loss, and the difference between the theoretical and actual safety indicators is taken as the negative impact of deterioration of safety margin. Based on the type label of the transformer area, the corresponding negative impact tolerance threshold is retrieved. When the negative impact of economic loss or the negative impact of deterioration of security margin exceeds the negative impact tolerance threshold, an overzealous preemption strategy alarm is generated.
[0173] This also includes: The frequency of daily mutual exclusion time intervals, the preemption ratio of each scenario, the negative impact of economic losses, and the negative impact of deteriorated security margins are packaged and uploaded to the cloud platform. The ratio of the difference between the actual operating deviation of each scenario after compensation and correction and the theoretical operating deviation before compensation and correction to the theoretical operating deviation is calculated as the improvement rate. When the improvement rate is lower than the preset improvement lower limit, it is determined that the current risk sensitivity coefficient is too low, and a risk sensitivity coefficient correction command is generated and sent to the local control terminal of the energy storage in the distribution area to increase the initial risk sensitivity coefficient.
[0174] The energy gap compensation module 14 is used to extract the unmet energy gap of the preempted scenario, and to add the unmet energy gap to the theoretical charging and discharging demand sequence of the subsequent non-mutually exclusive period to compensate and correct the multi-scenario allocation scheme affected by competition.
[0175] This includes extracting the unmet energy gap in the preempted scenario, including: The difference between the theoretical charge or discharge energy and the actual energy executed in scenarios where power allocation is deprived or discharge depth is limited within the mutually exclusive timing interval is calculated as the unmet energy gap. The unmet energy gaps are marked with scene tags and timestamps and stored in the competition compensation queue.
[0176] The unmet energy gap is rolled over and accumulated into the theoretical charge / discharge demand sequence of subsequent non-mutually exclusive periods to compensate and correct the multi-scenario allocation scheme affected by competition, including: In subsequent non-mutually exclusive periods, when there is a margin of available energy storage power, unmet energy gaps are extracted from the competition compensation queue in chronological order of timestamps and added to the theoretical charging and discharging demand sequence of the current period to perform compensation charging and discharging. When the compensated charging and discharging conflict with the theoretical charging and discharging demand for the current period, the demand for the current period takes priority, and the uncompensated gap remains in the competitive compensation queue to wait for the next compensation opportunity. The failure period is matched according to the scenario label of the unmet energy gap, wherein the first failure period corresponding to the emergency backup scenario is shorter than the second failure period corresponding to the peak shaving scenario. When the unmet energy gap remains in the competitive compensation queue for longer than the corresponding failure period, it is removed from the queue, and compensation correction is stopped.
[0177] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.
[0178] This specification and accompanying drawings are merely illustrative examples of the invention and are intended to cover any and all modifications, variations, combinations, or equivalents within the scope of the invention. Clearly, those skilled in the art can make various alterations and modifications to the invention without departing from its scope. Therefore, if such modifications and modifications fall within the scope of the invention and its equivalents, the invention is intended to include these modifications and modifications.
Claims
1. A method for configuring energy storage in transformer substations under multi-scenario competition, characterized in that, include: The theoretical charging and discharging demand sequence of each application scenario of energy storage in the transformer area is obtained, and based on the power direction and target state of charge of the theoretical charging and discharging demand sequence in time sequence, the mutually exclusive time sequence intervals of multi-scenario competition are identified. Collect real-time operation data of the station area, calculate the real-time operation deviation between the real-time operation data and the preset safety boundary of each scenario, and dynamically correct the scenario preemption weight of each scenario in the mutually exclusive time interval based on the nonlinear mapping relationship of the real-time operation deviation. Within the mutually exclusive time interval, based on the corrected scenario preemption weight, the available power and available capacity of the energy storage in the transformer area are preemptively allocated to generate the actual charging and discharging configuration curve under multi-scenario competition. Extract the unmet energy gaps in the preempted scenarios, and add these unmet energy gaps to the theoretical charging and discharging demand sequence in subsequent non-mutually exclusive periods to compensate and correct the multi-scenario allocation scheme affected by competition.
2. The method for configuring energy storage in a transformer substation under multi-scenario competition as described in claim 1, characterized in that, Obtain the theoretical charge and discharge demand sequence for each application scenario of energy storage in the distribution area, and based on the power direction and target state of charge of the theoretical charge and discharge demand sequence in time sequence, identify the mutually exclusive time sequence intervals of multi-scenario competition, including: The theoretical charge and discharge demand sequence for peak shaving scenarios, new energy consumption scenarios, dynamic capacity expansion scenarios, and emergency backup scenarios is loaded. When any scenario requires charging and another scenario requires discharging under the same time sequence, it is determined that power direction mutual exclusion has occurred and marked as a power mutual exclusion time sequence interval. When any scenario in the same time series requires the stored energy to be released to the first target state of charge, and another scenario requires the stored energy to be retained to the second target state of charge, which is greater than the first target state of charge, it is determined that a mutual exclusion of state of charge has occurred and is marked as an energy mutual exclusion time series interval.
3. The method for configuring energy storage in a transformer substation under multi-scenario competition as described in claim 1, characterized in that, Collect real-time operational data of the control area, and calculate the real-time operational deviation between the real-time operational data and the preset safety boundaries for each scenario, including: Real-time load rate of transformers, real-time output value of photovoltaic power, and real-time voltage drop frequency of the power grid are collected from the measurement terminal in the distribution area. The difference between the real-time load rate of the transformer and the preset capacity expansion trigger boundary is calculated as the operating deviation of the dynamic capacity expansion scenario; The difference between the real-time photovoltaic output value and the theoretically predicted output curve is calculated as the operational deviation of the new energy consumption scenario; The difference between the real-time voltage drop frequency of the power grid and the historical normal reference frequency is calculated as the operational deviation of the emergency backup scenario.
4. The method for configuring energy storage in a transformer substation under multi-scenario competition as described in claim 3, characterized in that, Based on the nonlinear mapping relationship of the real-time operational deviation, the scene preemption weight of each scene within the mutually exclusive time interval is dynamically adjusted, including: An exponential preemption weight function is constructed, which takes real-time running deviation as input and scene preemption weight as output. The scene preemption weight increases exponentially with the increase of real-time running deviation. Obtain the transformer area type label, which is used to identify the electricity consumption characteristic type of the transformer area, including at least one of industrial transformer areas, commercial transformer areas and residential transformer areas; The initial risk sensitivity coefficient is matched according to the type label of the transformer area. The initial risk sensitivity coefficient is used to control the response rate of the scene preemption weight as the real-time operation deviation increases. The frequency of occurrence of historical mutually exclusive time series intervals within a preset time window is counted. When the frequency of occurrence exceeds a preset fluctuation threshold, the initial risk sensitivity coefficient is amplified according to the ratio of the frequency of occurrence to the preset fluctuation threshold, and the amplified risk sensitivity coefficient does not exceed a preset upper limit value, to obtain the current risk sensitivity coefficient. The real-time operational deviation of each scenario and the current risk sensitivity coefficient are input into the exponential preemption weight function to calculate the dynamic preemption weight of each scenario in the current time series.
5. The method for configuring energy storage in a transformer substation under multi-scenario competition as described in claim 1, characterized in that, Within the mutually exclusive time interval, based on the corrected scenario preemption weights, a preemptive allocation is performed on the available power and available capacity of the energy storage in the transformer area, generating actual charge-discharge configuration curves under multi-scenario competition, including: When in the power mutual exclusion time interval, compare the dynamic preemption weights of each scenario, allocate the total available power of the energy storage converter in the distribution area to the scenario with the largest preemption weight, and suspend the power allocation of the other competing scenarios. When in the energy mutual exclusion time series interval, the preemption weights of peak shaving scenario and emergency backup scenario are compared. If the weight of peak shaving scenario is dominant, energy storage is allowed to discharge to the first target state of charge. If the weight of emergency backup scenario is dominant, the energy storage discharge depth is forcibly limited, and the bottom line of state of charge is raised to the second target state of charge. Based on the full allocation results and the rise in the state of charge baseline, the actual charge-discharge configuration curve is generated.
6. The method for configuring energy storage in a transformer substation under multi-scenario competition as described in claim 5, characterized in that, Extract the unmet energy gap in the preempted scenario, including: The difference between the theoretical charge or discharge energy and the actual energy executed in scenarios where power allocation is deprived or discharge depth is limited within the mutually exclusive timing interval is calculated as the unmet energy gap. The unmet energy gaps are marked with scene tags and timestamps and stored in the competition compensation queue.
7. The method for configuring energy storage in a transformer substation under multi-scenario competition as described in claim 6, characterized in that, The unmet energy gap is rolled over and accumulated into the theoretical charge / discharge demand sequence of subsequent non-mutually exclusive periods to compensate and correct the multi-scenario allocation scheme affected by competition, including: In subsequent non-mutually exclusive periods, when there is a margin of available energy storage power, unmet energy gaps are extracted from the competition compensation queue in chronological order of timestamps and added to the theoretical charging and discharging demand sequence of the current period to perform compensation charging and discharging. When the compensated charging and discharging conflict with the theoretical charging and discharging demand for the current period, the demand for the current period takes priority, and the uncompensated gap remains in the competitive compensation queue to wait for the next compensation opportunity. The failure period is matched according to the scenario label of the unmet energy gap, wherein the first failure period corresponding to the emergency backup scenario is shorter than the second failure period corresponding to the peak shaving scenario. When the unmet energy gap remains in the competitive compensation queue for longer than the corresponding failure period, it is removed from the queue, and compensation correction is stopped.
8. The method for configuring energy storage in a transformer substation under multi-scenario competition as described in claim 1, characterized in that, After generating actual charge / discharge configuration curves under multi-scenario competition, the following is also included: Obtain the theoretical benefit index and theoretical safety index of the preempted scenario within the mutually exclusive time interval. The theoretical benefit index is the expected benefit value calculated based on the theoretical charging and discharging requirements of each scenario. The theoretical safety index is the expected safety margin value calculated based on the preset safety boundary of each scenario. Calculate the actual benefit index and actual safety index after the preemption. The actual benefit index is the actual benefit value calculated based on the actual charge and discharge configuration curve, and the actual safety index is the actual safety margin value calculated based on the actual charge and discharge configuration curve. The difference between the theoretical and actual return indicators is taken as the negative impact of economic loss, and the difference between the theoretical and actual safety indicators is taken as the negative impact of deterioration of safety margin. Based on the type label of the transformer area, the corresponding negative impact tolerance threshold is retrieved. When the negative impact of economic loss or the negative impact of deterioration of security margin exceeds the negative impact tolerance threshold, an overzealous preemption strategy alarm is generated.
9. The method for configuring energy storage in a transformer substation under multi-scenario competition as described in claim 8, characterized in that, Also includes: The frequency of daily mutual exclusion time intervals, the preemption ratio of each scenario, the negative impact of economic losses, and the negative impact of deteriorated security margins are packaged and uploaded to the cloud platform. The ratio of the difference between the actual operating deviation of each scenario after compensation and correction and the theoretical operating deviation before compensation and correction to the theoretical operating deviation is calculated as the improvement rate. When the improvement rate is lower than the preset improvement lower limit, it is determined that the current risk sensitivity coefficient is too low, and a risk sensitivity coefficient correction command is generated and sent to the local control terminal of the energy storage in the distribution area to increase the initial risk sensitivity coefficient.
10. A transformer substation energy storage configuration system under multi-scenario competition, characterized in that, A method for configuring transformer substation energy storage under multi-scenario competition as described in any one of claims 1 to 9, comprising: The mutual exclusion timing identification module is used to obtain the theoretical charging and discharging demand sequence of each application scenario of energy storage in the distribution area, and to identify the mutually exclusive timing intervals of multiple scenarios based on the power direction and target state of charge of the theoretical charging and discharging demand sequence in the timing. The preemption weight correction module is used to collect real-time operation data of the transformer area, calculate the real-time operation deviation between the real-time operation data and the preset safety boundary of each scenario, and dynamically correct the scenario preemption weight of each scenario within the mutually exclusive time interval according to the nonlinear mapping relationship of the real-time operation deviation. The preemptive allocation module is used to perform preemptive allocation of the available power and available capacity of the energy storage in the transformer area based on the corrected scenario preemption weight within the mutually exclusive time interval, and generate the actual charging and discharging configuration curve under multi-scenario competition. The energy gap compensation module is used to extract the unmet energy gaps in the preempted scenarios and add the unmet energy gaps to the theoretical charging and discharging demand sequence of subsequent non-mutually exclusive periods to compensate and correct the multi-scenario allocation scheme affected by competition.
