A method and system for coordinated energy scheduling of optical storage based on net load prediction

Abstract

This invention discloses a photovoltaic-storage coordinated energy dispatching method and system based on net load forecasting, belonging to the field of energy management and power dispatching technology. The method includes: acquiring the predicted load value of the power supply zone, the predicted photovoltaic output value, and the energy storage operating status; generating a net load forecast interval; and identifying the photovoltaic absorption window, the demand shaving window, and the uncertain reversal window accordingly; generating a dynamic adjustment corridor for the energy storage state of charge based on the charging space required by the photovoltaic absorption window and the discharging capacity required by the demand shaving window; generating an energy storage charging and discharging plan based on the energy storage operating status, and performing rolling corrections based on the actual net load deviation; issuing energy storage power commands and receiving execution feedback. This invention can reduce the risk of premature charging or depletion of energy storage caused by forecasting errors, and improve photovoltaic absorption capacity and demand shaving effect.

Description

Technical Field

[0001] This invention relates to the field of energy management and power dispatching technology, and in particular to a photovoltaic-storage coordinated energy dispatching method and system based on net load forecasting. Background Technology

[0002] With the integration of distributed photovoltaic (PV) power, electrochemical energy storage, charging piles, and flexible loads into the power supply systems of rail transit, industrial parks, and large public buildings, traditional power supply zones, which are mainly based on unidirectional power supply and passive power consumption, are gradually transforming into comprehensive energy units with the capabilities of power generation, energy storage, energy consumption regulation, and cross-regional coordination. In such scenarios, PV output is greatly affected by weather, irradiance, and cloud cover changes, while the load of power supply zones fluctuates significantly due to passenger flow, operating hours, air conditioning load, and equipment operating status. If PV output and load demand cannot be effectively predicted and energy storage charging and discharging cannot be rationally arranged, it will easily lead to problems such as the inability to absorb surplus PV power, excessive grid power consumption during peak electricity demand periods, and insufficient utilization efficiency of energy storage resources.

[0003] Existing photovoltaic-storage coordinated dispatch methods typically control energy storage charging and discharging based on fixed time-of-use pricing periods, the current surplus state of photovoltaic power, or the current peak load state. That is, energy storage charging is scheduled during off-peak pricing periods or when photovoltaic power is surplus, and energy storage discharging is scheduled during peak pricing periods or when load is peak. Although this method has a simple control logic, it focuses more on the current period or the preset peak-valley period and lacks a unified judgment on the relationship between future net load changes, subsequent photovoltaic absorption demand, and subsequent peak shaving demand. When the actual photovoltaic output and the load of the power supply area deviate from the forecast results, the energy storage may be fully charged before the subsequent large photovoltaic power generation, resulting in the surplus photovoltaic power being unable to continue to absorb power; or it may be over-discharged before the actual demand peak arrives, resulting in insufficient subsequent peak shaving capacity. Summary of the Invention

[0004] The purpose of this invention is to address the shortcomings of existing technologies where, when there are prediction deviations between the load of power supply zones and photovoltaic output, energy storage is easily filled or depleted in advance, leading to insufficient subsequent photovoltaic absorption capacity and insufficient demand peak shaving capacity. Therefore, this invention proposes a photovoltaic-storage coordinated energy dispatching method and system based on net load prediction.

[0005] To address the problems existing in the prior art, the present invention adopts the following technical solution: A photovoltaic-storage coordinated energy dispatch method based on net load forecasting includes: S1. Obtain power supply zone prediction data and energy storage operation status, and generate net load prediction intervals based on the power supply zone prediction data; S2. Identify key scheduling windows based on the net load forecast interval. The key scheduling windows include photovoltaic consumption window, demand shaving window and uncertain reversal window. S3. Generate a dynamic adjustment corridor for the energy storage state of charge (ESC) to limit the upper and lower limits of the ESC operation based on the photovoltaic absorption window and the demand peak shaving window. S4. Generate an energy storage charging and discharging plan based on the energy storage state of charge, the dynamic adjustment corridor, the net load prediction range, and the energy storage operation status. S5. Obtain the actual net load deviation, and perform rolling correction on the energy storage charging and discharging plan based on the actual net load deviation to obtain the corrected energy storage charging and discharging plan. S6. Convert the energy storage charging and discharging plan or the modified energy storage charging and discharging plan into an energy storage power command.

[0006] Preferably, the power supply zone prediction data includes the power supply zone load prediction value, photovoltaic output prediction value, load prediction confidence interval, photovoltaic prediction confidence interval, time-of-use electricity price information, and demand control target; the energy storage operation status includes the energy storage state of charge, energy storage rated capacity, energy storage maximum charging power, energy storage maximum discharging power, charging efficiency, discharging efficiency, and energy storage allowable state of charge range.

[0007] Preferably, generating a net load forecast interval based on the power supply zone forecast data includes: The net load forecast value is generated based on the difference between the predicted load value of the power supply zone and the predicted photovoltaic output value, and the net load forecast interval is generated based on the load forecast confidence interval and the photovoltaic forecast confidence interval. The net load forecast interval includes a lower bound and an upper bound. The lower bound is determined by the lower bound of the load forecast and the upper bound of the photovoltaic forecast, and the upper bound is determined by the upper bound of the load forecast and the lower bound of the photovoltaic forecast.

[0008] Preferably, identifying key scheduling windows based on the net load forecast interval includes: When the upper limit of the net load forecast is less than zero, the corresponding time period will be marked as the photovoltaic consumption window; When the upper limit of the net load forecast is greater than the demand control target, the corresponding time period will be marked as the demand shaving window. When the lower bound of net load forecast is less than zero and the upper bound of net load forecast is greater than zero, the corresponding time period is marked as an uncertain inversion window.

[0009] Preferably, a dynamic adjustment corridor for the energy storage state of charge (SBC) is generated based on the photovoltaic absorption window and the demand peak shaving window to limit the upper and lower limits of the SBC operation, including: The amount of rechargeable space that should be reserved by energy storage before entering the photovoltaic absorption window is determined based on the surplus photovoltaic power within the photovoltaic absorption window. The amount of dischargeable energy that the energy storage should retain before entering the demand shaving window is determined based on the power demand exceeding the demand control target within the demand shaving window. A dynamic adjustment corridor for the energy storage state of charge is generated based on the rechargeable space, the dischargeable capacity, and the allowable state of charge range.

[0010] Preferably, when the lower limit of the energy storage state of charge dynamically adjusting corridor based on the rechargeable space and the dischargeable power is higher than the upper limit, the dischargeable power required by the demand peak shaving window is retained, and the unmet photovoltaic power consumption is output to the upper-level scheduling system as a quantity to be coordinated.

[0011] Preferably, the energy storage charging and discharging plan is generated based on the energy storage state of charge dynamic adjustment corridor, net load prediction range, and energy storage operating status, including: When the current time period falls within the photovoltaic consumption window and the energy storage state of charge is below the upper limit of the energy storage state of charge dynamic adjustment corridor, an energy storage charging plan is generated. When the current time period falls within the demand shaving window and the energy storage state of charge is higher than the lower limit of the energy storage state of charge dynamic adjustment corridor, an energy storage discharge plan is generated. When the current time period is neither a photovoltaic consumption window nor a demand peak shaving window, a charging plan or discharging plan is generated to bring the energy storage state of charge back into the energy storage state of charge dynamic adjustment corridor, or a standby plan is generated to keep the energy storage in the energy storage state of charge dynamic adjustment corridor, based on the position of the energy storage state of charge relative to the energy storage state of charge dynamic adjustment corridor.

[0012] Preferably, the actual net load deviation is obtained, and the energy storage charging and discharging plan is rolled over and corrected based on the actual net load deviation, including: The actual load, actual photovoltaic output, and actual state of charge of energy storage in the power supply zone are obtained, and the actual net load is generated based on the difference between the actual load and the actual photovoltaic output of the power supply zone. The actual net load deviation is generated based on the difference between the actual net load and the predicted net load. When the actual net load deviates continuously from the net load prediction range within the sampling sequence corresponding to the rolling correction cycle, or when the actual net load within the uncertain reversal window has been determined to be in a positive or negative state, the energy storage charging and discharging plan for the subsequent time period is regenerated based on the latest actual net load and the actual state of charge of the energy storage.

[0013] Preferably, converting the energy storage charge / discharge plan or the modified energy storage charge / discharge plan into an energy storage power command includes: The energy storage charging and discharging plan or the revised energy storage charging and discharging plan will be sent to the energy storage energy management system or the energy storage converter. Obtain the actual charging and discharging power, energy storage state of charge, and equipment operating status fed back by the energy storage management system or energy storage converter; When the feedback result indicates that the energy storage cannot execute the energy storage power command, the executable energy storage power command for the current time period is re-determined based on the actual available energy storage power, and the unmet photovoltaic power consumption or peak shaving power gap is output to the upper-level dispatch system.

[0014] Compared with the prior art, the beneficial effects of the present invention are: 1. This invention generates a net load prediction interval by acquiring the predicted load value of the power supply zone, the predicted photovoltaic output value, and the corresponding prediction confidence interval. Based on the net load prediction interval, it identifies the photovoltaic absorption window, the demand shaving window, and the uncertain reversal window. This enables energy storage dispatch to no longer rely solely on fixed peak-valley electricity price periods or the current single-point net load status for charging and discharging. Instead, it can predict in advance whether there is a stable photovoltaic surplus, demand exceeding the limit risk, or net load reversal risk in subsequent periods. In the event of prediction deviation, it can expand the energy storage regulation target from power balance in a single period to a continuous regulation process oriented towards subsequent critical dispatch windows. This avoids the energy storage being prematurely charged before the photovoltaic power generation peak or prematurely depleted before the actual demand peak arrives, thereby improving the photovoltaic absorption capacity and the reliability of demand shaving.

[0015] 2. This invention further generates a dynamic adjustment corridor for the energy storage state of charge based on the rechargeable space required by the photovoltaic absorption window and the dischargeable capacity required by the demand peak shaving window. This dynamic adjustment corridor constrains the current charging, discharging, or standby actions of the energy storage, ensuring that the energy storage state of charge, while meeting the safe operating range of the equipment, retains the adjustment capability required for subsequent photovoltaic absorption and demand reduction. The actual net load deviation is used to perform rolling corrections on the energy storage charging and discharging plan, and the executable energy storage power command is re-determined in conjunction with energy storage execution feedback, forming a closed-loop control of day-ahead planning, intraday rolling, and real-time correction. This can reduce the risks of ineffective cycle charging and discharging, insufficient peak shaving power, and surplus photovoltaic power transmission caused by the accumulation of prediction errors, and improve the stability of energy storage operation, dispatch economy, and the effect of photovoltaic-energy storage coordinated operation in power supply zones. Attached Figure Description

[0016] The accompanying drawings, which are included to provide a further understanding of the invention and form part of this application, illustrate exemplary embodiments of the invention and, together with their description, serve to explain the invention and do not constitute an undue limitation thereof. In the drawings: Figure 1 This is a flowchart illustrating a photovoltaic-storage coordinated energy dispatching method based on net load prediction, provided as an embodiment of the present invention. Detailed Implementation

[0017] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments.

[0018] Example 1, as Figure 1 As shown, this embodiment provides a photovoltaic-storage coordinated energy dispatching method based on net load forecasting, applied to power supply zones equipped with distributed photovoltaic, electrochemical energy storage, and energy management platforms; the power supply zone can be a subway depot power supply zone, a station group power supply zone, a switchgear power supply range, a main substation power supply range, or other energy-consuming areas with photovoltaic power generation and energy storage regulation capabilities; the energy management platform is communicatively connected to the load metering system, photovoltaic monitoring system, energy storage management system, and energy storage converter, used to acquire forecast data, operating status, and actual measurement data, and to issue energy storage power commands to the energy storage management system or energy storage converter; In this embodiment, a power supply zone refers to an area covered by the same power supply node or the same power supply management boundary; net load refers to the difference between the load of the power supply zone and the photovoltaic output. A positive net load indicates that the power supply zone needs to draw power from the grid, while a negative net load indicates that the power supply zone has photovoltaic surplus; energy storage state of charge refers to the ratio of the current available energy storage capacity to the rated capacity of the energy storage; the energy storage state of charge dynamic adjustment corridor refers to the upper and lower limits of the energy storage state of charge operation dynamically generated within the allowable state of charge range of energy storage, based on the charging space required by the future photovoltaic consumption window and the discharging capacity required by the future demand peak shaving window.

[0019] The method in this embodiment includes the following steps; S1. Obtain power supply zone prediction data and energy storage operation status, and generate net load prediction intervals based on power supply zone prediction data; Specifically, the energy management platform uses power supply zones as scheduling units and obtains power supply zone prediction data and energy storage operation status within future scheduling cycles according to scheduling time periods. The scheduling time periods are consistent with the prediction granularity, scheduling granularity, or metering granularity of the energy management platform, for example, using fifteen minutes as a scheduling time period. The future scheduling cycle is the 24-hour cycle corresponding to the current day's plan, or a future several-hour cycle corresponding to intraday rolling correction. The power supply zone forecast data includes the power supply zone load forecast, photovoltaic output forecast, load forecast confidence interval, photovoltaic forecast confidence interval, time-of-use pricing information, and demand control targets. The power supply zone load forecast is generated by the energy management platform based on historical load data, passenger flow data, meteorological data, and operating schedules. The photovoltaic output forecast is generated by the energy management platform based on historical power generation data, meteorological forecast data, photovoltaic module installation parameters, and time stamps. The time-of-use pricing information is used to determine whether the current time period is a low-valley, flat, peak, or peak electricity consumption period. The demand control targets are the demand declaration values ​​for the power supply zone, or the demand control values ​​formed by the energy management platform based on the monthly maximum demand management requirements. The load prediction confidence interval and the photovoltaic prediction confidence interval are directly output by the corresponding prediction models, or obtained by correcting the predicted values ​​based on the historical prediction error distribution. Specifically, after generating the load prediction value for a power supply zone, the energy management platform obtains the historical load prediction error of that power supply zone at the same time granularity, and determines the upper and lower bounds of the load prediction according to the positive and negative deviations of the historical load prediction error. After generating the photovoltaic output prediction value, the energy management platform obtains the historical photovoltaic prediction error of the corresponding photovoltaic site or power supply zone aggregated photovoltaic at the same time granularity, and determines the upper and lower bounds of the photovoltaic prediction according to the positive and negative deviations of the historical photovoltaic prediction error. The load prediction confidence interval is used to represent the prediction range that the load of the power supply zone may fall into within the same time period, and the photovoltaic prediction confidence interval is used to represent the prediction range that the photovoltaic output may fall into within the same time period. When using the historical prediction error distribution to form the prediction confidence interval, the historical prediction error comes from the difference between the predicted value and the actual measured value already saved by the energy management platform, and no new independent acquisition equipment is required. The energy storage operating status includes the energy storage state of charge, rated energy storage capacity, maximum energy storage charging power, maximum energy storage discharging power, charging efficiency, discharging efficiency, and allowable state of charge range. The energy storage state of charge, actual charging and discharging power, and equipment availability are fed back by the energy storage management system or energy storage converter. The allowable state of charge range is determined by the safety operation requirements of the energy storage equipment, including the lower limit and upper limit of the allowable state of charge.

[0020] Within each scheduling period, the energy management platform generates a net load forecast based on the difference between the predicted load for the power supply zone and the predicted photovoltaic output. ; in, This represents the net load forecast for time period t. This represents the predicted load value for the power supply zone during time period t. This represents the predicted photovoltaic power output for time period t. like A value greater than zero indicates that the power supply zone needs to draw power from the mains during time period t; if A value less than zero indicates that the power supply zone has a photovoltaic surplus during time period t; if A value of zero indicates that the predicted load and photovoltaic output of the power supply zone have reached a predicted balance during this period. Furthermore, the energy management platform generates a net load forecast interval based on the load forecast confidence interval and the photovoltaic forecast confidence interval; if the load forecast confidence interval for time period t is... The confidence interval for photovoltaic prediction is: Then the net load forecast interval includes the lower bound of the net load forecast. and the upper limit of net load forecast And determine it in the following way: ; ; in, This is the lower bound for load forecasting over time period t. This is the upper bound for load forecasting over time period t. This is the lower bound of the photovoltaic prediction for time period t. The upper bound of the photovoltaic forecast is the time period t. The lower bound of the net load forecast is calculated by using the lower bound of the load forecast and the upper bound of the photovoltaic forecast to reflect the maximum photovoltaic surplus state that may occur when the load is low and the photovoltaic power is high. The upper bound of the net load forecast is calculated by using the upper bound of the load forecast and the lower bound of the photovoltaic forecast to reflect the maximum power consumption state that may occur when the load is high and the photovoltaic power is low. After processing by S1, the energy management platform obtains the net load forecast value, the lower bound of the net load forecast, and the upper bound of the net load forecast for each scheduling time period within the future scheduling cycle; the net load forecast interval serves as the input for subsequent identification of key scheduling windows.

[0021] S2. Identify key scheduling windows based on the net load forecast interval. The key scheduling windows include the photovoltaic consumption window, the demand peak shaving window, and the uncertain reversal window. Specifically, the energy management platform determines the net load forecast interval for each scheduling period within the future scheduling cycle and assigns a window identifier to the scheduling period based on the determination result. When time interval t satisfies: ; This indicates that even if calculated according to the upper limit of the net load forecast, the power supply zone is still in a state of photovoltaic surplus during this period; at this time, the period t is marked as the photovoltaic absorption window; the photovoltaic absorption window indicates that energy storage needs to reserve rechargeable space before entering this period in order to absorb the surplus photovoltaic power and reduce the demand for curtailment or external transmission. When time interval t satisfies: ; This indicates that, based on the upper limit of the net load forecast, the power supply zone faces the risk of exceeding the demand control target during this time period; therefore, time period t is marked as the demand peak shaving window; where, The demand control target is the demand peak shaving window, which indicates that the energy storage needs to retain dischargeable electricity before entering this time period so that it can discharge and shave off peaks when the demand approaches or exceeds the control target. When time interval t satisfies: ; This indicates that the net load forecast interval for this period crosses zero point. The power supply zone may switch between photovoltaic surplus state and grid power supply state due to load forecast deviation or photovoltaic forecast deviation. At this time, the time period t is marked as an uncertain reversal window. The uncertain reversal window is not used as the basis for immediate forced charging or immediate forced discharging, but as the key review object in intraday rolling correction and real-time correction. If the conditions for both the demand shaving window and the uncertain reversal window are met simultaneously in the same time period, the energy management platform will mark the time period as the demand shaving window and add an uncertain reversal identifier. When generating the dynamic adjustment corridor for energy storage state of charge, this time period will be included in the calculation according to the demand shaving window. During the subsequent rolling correction, this time period will still be the key object for review of the net load positive and negative states. After S2 processing, the energy management platform obtains the photovoltaic consumption window set, the demand shaving window set, and the uncertain reversal window set for the future scheduling cycle; the photovoltaic consumption window set and the demand shaving window set are used to generate the energy storage state of charge dynamic adjustment corridor, and the uncertain reversal window set is used to trigger or assist intraday rolling correction.

[0022] S3. Generate a dynamic adjustment corridor for the energy storage state of charge (SOC) to limit the upper and lower limits of the energy storage SOC operation based on the photovoltaic absorption window and the demand peak shaving window. Specifically, the energy management platform uses the current state of charge of the energy storage as its initial state, and calculates the rechargeable space and dischargeable capacity that the energy storage needs to retain at each time point in future scheduling cycles in chronological order; the change in the state of charge of the energy storage satisfies the following energy conservation relationship: ; in, The energy storage state of charge during time period t. To schedule the time step, For energy storage charging power, E represents the energy storage discharge power, and E represents the rated energy storage capacity. For charging efficiency, For discharge efficiency; this relationship indicates that the state of charge during energy storage charging increases with the energy input, and the state of charge during energy storage discharging decreases with the energy released. For any time The energy management platform statistics are from time to time The photovoltaic (PV) grid connection window from the beginning to the end of the scheduling cycle forms a set of PV grid connection windows. For the photovoltaic grid integration window set For each time period, the energy management platform determines the amount of surplus photovoltaic power that needs to be absorbed by energy storage based on the stable surplus photovoltaic power corresponding to a net load forecast upper bound that is less than zero, the maximum charging power of energy storage, and the dispatch time step. Considering charging efficiency, the energy storage side should reserve rechargeable space for subsequent photovoltaic consumption. Determine as follows: ; in, This refers to the rechargeable storage space required for the photovoltaic power generation window after time T. This represents the maximum charging power for energy storage; [The following is a list of parameters:] ...using... The reason for including it in the calculation is that when the upper limit of the net load forecast is still less than zero, this period still has a stable photovoltaic surplus after taking into account the forecast deviation, which is suitable as the basis for reserving charging space for energy storage. For any given time T, the energy management platform statistically analyzes the demand shaving windows from time T to the end of the scheduling cycle, forming a set of demand shaving windows. For demand shaving window set For each time period, the energy management platform determines the peak-shaving power that needs to be released by energy storage for that period based on the power exceeding the demand control target by the upper limit of the net load forecast, the maximum discharge power of energy storage, and the scheduling time step. After considering discharge efficiency, the energy storage side determines the dischargeable power that should be retained for subsequent demand peak shaving. Determine as follows: ; in, This refers to the amount of dischargeable energy stored on the energy storage side required for the peak shaving window after time T. This represents the maximum discharge power of the energy storage. The energy management platform is based on the rechargeable space. Dischargeable capacity The allowable state of charge range of energy storage generates a dynamic adjustment corridor for the energy storage state of charge; if the allowable state of charge range of energy storage is... Then the dynamic lower limit of the energy storage state of charge at time T. and the dynamic upper limit of the state of charge of energy storage Determine as follows: ; ; in, Used to reserve dischargeable capacity for subsequent demand shaving windows. This is used to reserve rechargeable space for subsequent photovoltaic power consumption; the energy storage state of charge dynamic adjustment corridor is... ; when Less than or equal to At time T, it indicates that energy storage has a feasible state-of-charge operating range, capable of simultaneously accommodating photovoltaic power consumption and demand peak shaving needs in subsequent dispatch cycles; at this time, the energy management platform will As a dynamic adjustment corridor for the state of charge of energy storage at time T; when Higher than When this occurs, it indicates that the energy storage capacity or power capability is insufficient to simultaneously meet the charging space required for the subsequent photovoltaic consumption window and the discharging capacity required for the demand peak shaving window; the energy management platform marks this moment as an energy storage regulation conflict state, and uses... The peak shaving and retention boundary is used to generate subsequent energy storage charging and discharging plans; in the event of energy storage regulation conflict, the energy management platform restricts the energy storage from continuing to discharge within the non-demand peak shaving window, so that the energy storage state of charge is not lower than the peak shaving and retention boundary; for the photovoltaic absorption window, the energy management platform generates an executable charging plan based on the actual chargeable space of the energy storage and the maximum charging power of the energy storage, and the excess photovoltaic power that the energy storage cannot absorb forms a quantity to be coordinated and outputs it to the upper-level dispatch system; after receiving the quantity to be coordinated, the upper-level dispatch system calls at least one of the following for coordinated absorption: orderly charging of charging piles, flexible load of air conditioning, or cross-regional mutual assistance resources, according to the adjustable resource status within the power supply zone or adjacent power supply zones; After S3 processing, the energy management platform obtains the dynamic adjustment corridor of energy storage charge status at each time point in the future scheduling cycle.

[0023] S4. Generate an energy storage charging and discharging plan by dynamically adjusting the corridor, net load prediction range and energy storage operation status based on the energy storage state of charge. Specifically, for each time period within a future scheduling cycle, the energy management platform determines the energy storage action type and energy storage power target by combining the key scheduling window type to which that time period belongs, the net load forecast, the current state of charge of energy storage, and the dynamic adjustment corridor of the state of charge of energy storage; the energy storage action type includes charging, discharging, and standby; When the current time period t falls within the photovoltaic consumption window, and the energy storage state of charge (SOC(t)) is lower than the upper limit of the energy storage state of charge dynamic adjustment corridor. At that time, the energy management platform generates an energy storage charging plan; energy storage charging power Determine as follows: ; in, Let t be the energy storage charging power during time period t; the above calculation ensures that the energy storage charging power does not exceed the photovoltaic surplus power corresponding to the net load forecast value, does not exceed the maximum energy storage charging power, and does not cause the energy storage state of charge to exceed the upper limit of the energy storage state of charge dynamic adjustment corridor. When the current time period t falls within the photovoltaic consumption window, but the energy storage state of charge (SOC(t)) has reached or exceeded the upper limit of the energy storage state of charge dynamic adjustment corridor. At that time, the energy management platform will no longer generate a charging plan to increase the state of charge of the energy storage. Instead, it will output the surplus photovoltaic power that may be generated during that period as a quantity to be coordinated to the upper-level dispatch system to avoid the inability to retain subsequent adjustment space after the energy storage is fully charged in advance. When the current time period t falls within the demand shaving window, and the energy storage state of charge (SOC(t)) is higher than the lower limit of the energy storage state of charge dynamic adjustment corridor. At that time, the energy management platform generates an energy storage discharge plan; the energy storage discharge power... Determine as follows: ; in, Let t be the energy storage discharge power during time period t; the above calculation ensures that the energy storage discharge power does not exceed the net load forecast value exceeding the demand control target, does not exceed the maximum energy storage discharge power, and does not cause the energy storage state of charge to fall below the lower limit of the energy storage state of charge dynamic adjustment corridor. When the current time period t falls within the demand shaving window, but the energy storage state of charge (SOC(t)) has reached or fallen below the lower limit of the energy storage state of charge dynamic adjustment corridor. At this time, the energy management platform will no longer generate a discharge plan to further reduce the state of charge of the energy storage, but will output the unmet peak shaving power gap to the upper-level dispatch system to avoid over-discharge of energy storage or affect the peak shaving capacity during subsequent critical periods. When the current time period t falls neither within the photovoltaic consumption window nor the demand peak shaving window, the energy management platform dynamically adjusts the corridor's position based on the energy storage state of charge (SOC) relative to the SOC, generating charging, discharging, or standby plans. If the SOC(t) is lower than... If the current time period falls within the off-peak period indicated by the time-of-use pricing information, or if the energy management platform receives instructions from the upper-level dispatch system regarding rechargeable power reserves, cross-regional mutual assistance input power, or flexible load collaborative consumption, then an energy storage charging plan will be generated to bring the energy storage charge state back within the dynamic adjustment corridor for the energy storage charge state; if Higher than If the current time period falls within the peak or peak period indicated by the time-of-use pricing information, or if the energy management platform receives a dischargeable power demand, cross-regional power output command, or power reduction command from the upper-level dispatch system, then an energy storage discharge plan is generated to bring the energy storage charge state back within the dynamic adjustment corridor of the energy storage charge state; if In If the system does not receive any instructions from the upper-level dispatch system regarding rechargeable power margin, dischargeable power demand, cross-regional mutual assistance input power instructions, cross-regional mutual assistance output power instructions, flexible load collaborative absorption instructions, or power reduction instructions, a standby plan is generated to keep the energy storage within the dynamic adjustment corridor of the energy storage charge state, and to prevent unrestrained arbitrage charging and discharging.

[0024] When generating an energy storage charging and discharging plan, the energy management platform also checks whether the charging and discharging power exceeds the maximum charging power or the maximum discharging power of the energy storage based on the energy storage operating status, and checks whether the energy storage state of charge after the plan is executed exceeds the allowable state of charge range of the energy storage. If the verification results do not meet the constraints of the energy storage operating status, the energy storage charging and discharging plan is truncated and corrected according to the actual available power of the energy storage and the allowable state of charge range of the energy storage. After processing by S4, the energy management platform generates an energy storage charging and discharging plan. The energy storage charging and discharging plan includes a time period number, action type, energy storage charging power, energy storage discharging power, target energy storage state of charge after plan execution, the type of key scheduling window to which it belongs, and quantities to be coordinated. The energy storage charging and discharging plan serves as the input for subsequent rolling corrections and power command issuance.

[0025] S5. Obtain the actual net load deviation, and make rolling corrections to the energy storage charging and discharging plan based on the actual net load deviation to obtain the corrected energy storage charging and discharging plan. Specifically, during the execution of the energy storage charging and discharging plan, the energy management platform obtains the actual load, actual photovoltaic output, and actual energy storage status of the power supply zone according to the daily rolling cycle and real-time correction cycle; the actual load of the power supply zone is provided by the load metering system or the power monitoring system, the actual photovoltaic output is provided by the photovoltaic monitoring system, and the actual energy storage status of the energy storage is fed back by the energy storage energy management system or the energy storage converter. The energy management platform generates the actual net load based on the difference between the actual load and the actual photovoltaic output of the power supply zone: ; in, The actual net load for time period t. The actual load of the power supply zone during time period t. The actual photovoltaic output during time period t; The energy management platform further generates the actual net load deviation based on the difference between the actual net load and the predicted net load: ; in, The actual net load deviation over time period t; When the actual net load continuously deviates from the net load prediction range within the sampling sequence corresponding to the rolling correction cycle, it indicates that the net load prediction status on which the original energy storage charging and discharging plan was based can no longer reflect the current operating status. The continuous deviation means that within a rolling correction cycle, the energy management platform obtains the actual net load of multiple valid sampling points in chronological order. If the actual net load of all the multiple valid sampling points is less than the lower limit of the net load prediction for the corresponding time period, or is greater than the upper limit of the net load prediction for the corresponding time period, then it is determined that the actual net load continuously deviates from the net load prediction range. If, from a certain valid sampling point to the end of the rolling correction cycle, the actual net load is continuously outside the net load prediction range for the corresponding time period, and the direction of deviation is consistent, it is also determined that the actual net load continuously deviates from the net load prediction range. The valid sampling point refers to the sampling point where the load metering system, photovoltaic monitoring system, and energy storage operating status all provide normal feedback. The energy management platform uses the latest actual net load, the latest power supply zone load prediction data, the latest photovoltaic output prediction data, and the actual energy storage state of charge as inputs to re-execute S1 to S4 and generate the corrected energy storage charging and discharging plan for subsequent time periods.

[0026] When the actual net load within the uncertain inversion window has been determined to be either positive or negative, the energy management platform reclassifies the uncertain inversion window based on the actual net load status. The positive or negative actual net load status is determined based on the valid sampled values ​​within the uncertain inversion window. When the uncertain inversion window has ended, the average actual net load of each valid sampled point within the window is used as the criterion. When the uncertain inversion window has not yet ended but requires rolling correction, the average actual net load of the currently acquired valid sampled points is used as the criterion. If the average actual net load is greater than zero, it is determined to be positive; if the average actual net load is negative... If the average load is less than zero, it is determined to be in a negative state. If the average actual net load is equal to zero, or if the effective sampled values ​​within the uncertain reversal window alternate between positive and negative values, the uncertain reversal window identifier is maintained, and the judgment continues in the next rolling correction cycle. If the actual net load is determined to be in a negative state, this time period will participate in the subsequent rolling correction according to the photovoltaic consumption window. If the actual net load is determined to be in a positive state, and there is a risk that the actual net load will exceed the demand control target, this time period will participate in the subsequent rolling correction according to the demand peak shaving window. If the actual net load is determined to be in a positive state but does not meet the demand peak shaving conditions, this time period will be treated as a normal power consumption period. When the actual net load is still within the net load forecast range, and the positive or negative state of the actual net load within the uncertain reversal window has not yet reached a stable result, the energy management platform does not regenerate a complete energy storage charging and discharging plan, but only fine-tunes the energy storage power target for the current time period; the fine-tuning is used to smooth out short-term fluctuations in photovoltaic power and sudden load changes, so that the exchange power between the power supply zone and the grid side remains stable; the fine-tuned energy storage power is still constrained by the maximum energy storage charging power, the maximum energy storage discharging power, the allowable state of charge range of energy storage, and the dynamic adjustment corridor of the state of charge of energy storage. After processing by S5, the energy management platform obtains a revised energy storage charging and discharging plan. The revised energy storage charging and discharging plan covers the time period after the current time period that has not yet been executed, and uses the actual state of charge of the energy storage as the new initial state to avoid cumulative deviation between the current day plan and the actual execution state.

[0027] S6. Convert the energy storage charging and discharging plan or the revised energy storage charging and discharging plan into energy storage power commands, and form a scheduling closed loop based on energy storage execution feedback; Specifically, the energy management platform converts the energy storage charging and discharging plan or the revised energy storage charging and discharging plan into an energy storage power instruction; the energy storage power instruction includes the instruction time, charging power target, discharging power target, target energy storage state of charge, and execution duration; for the same time period, the energy storage power instruction only allows one of the charging power target and the discharging power target to be a valid power value, and the other to be zero, so as to avoid the energy storage from charging and discharging at the same time. The energy management platform sends energy storage power commands to the energy storage management system or energy storage converter. After receiving the energy storage power commands, the energy storage management system or energy storage converter performs charging or discharging control according to the equipment operating status, energy storage state of charge, available capacity of the energy storage converter and safety protection conditions, and feeds back the actual charging and discharging power, energy storage state of charge and equipment operating status to the energy management platform. When the equipment operating status feedback from the energy storage management system or energy storage converter indicates that the energy storage can execute the energy storage power command, the energy management platform records the actual execution result for that time period and uses the actual charging and discharging power and energy storage state of charge as inputs for the scheduling calculation of the next time period, thus forming a closed loop of prediction, planning, execution and feedback. When the equipment operating status feedback from the energy storage management system or energy storage converter indicates that energy storage cannot be executed at the target power, the energy management platform redetermines the executable energy storage power for the current time period based on the feedback of the actual available charging power, actual available discharging power, actual state of charge, and allowable state of charge range. If the current time period is for charging, the executable energy storage power shall not exceed the actual available charging power, and the state of charge after execution shall not exceed the upper limit of the allowable state of charge or the upper limit of the dynamic adjustment corridor for the state of charge. If the current time period is for discharging, the executable energy storage power shall not exceed the actual available discharging power, and the state of charge after execution shall not be lower than the lower limit of the allowable state of charge or the lower limit of the dynamic adjustment corridor for the state of charge. If the redetermined executable energy storage power is zero, the energy storage action for that time period shall be adjusted to standby, and the corresponding unmet photovoltaic power consumption or peak shaving power gap shall be output to the upper-level dispatch system. When feedback indicates that the energy storage cannot execute the energy storage power command due to fault, power limitation, limited energy storage state of charge, communication abnormality, or equipment protection action, the energy management platform re-determines the executable energy storage power command for the current time period based on the actual available energy storage power. For the portion that cannot be executed by energy storage, if it falls within the photovoltaic consumption window, the unmet photovoltaic consumption power will be output to the upper-level dispatch system as a quantity to be coordinated; if it falls within the demand peak shaving window, the unmet peak shaving power gap will be output to the upper-level dispatch system. The upper-level dispatch system determines whether to call up the orderly charging of charging piles and flexible air conditioning loads within the power supply zone, or coordinate the adjustable resources of adjacent power supply zones through cross-regional mutual assistance, based on the quantity to be coordinated or the peak shaving power gap.

[0028] Example 2: This example, based on Example 1, explains the method for generating power supply zone prediction data; The power supply zone load forecast is generated by the energy management platform based on historical load data, passenger flow data, meteorological data, and operating timetables. Historical load data refers to the total power output of the power supply zone over a historical period. Passenger flow data includes the inbound and outbound passenger flow or passenger flow forecasts for stations within the power supply zone. Meteorological data includes outdoor temperature, humidity, and other meteorological factors affecting air conditioning load. The operating timetable includes operating hours, shutdown hours, first and last train times, weekday indicators, and holiday indicators. The energy management platform uses a time series forecasting model to output the power supply zone load forecast and its confidence interval. The photovoltaic (PV) output forecast is generated by the energy management platform based on historical power generation data of PV sites, weather forecast data, PV module installation parameters, and time stamps. The weather forecast data includes irradiance, ambient temperature, relative humidity, and cloud cover. The PV module installation parameters include PV module tilt angle, azimuth angle, and rated installed capacity. The energy management platform first generates a PV output baseline value based on the theoretical power generation curve under clear sky conditions, and then performs residual correction on the PV output baseline value based on meteorological factors to obtain the PV output forecast value and the PV forecast confidence interval. In this embodiment, the predicted load value of the power supply zone, the predicted photovoltaic output value, the load prediction confidence interval, and the photovoltaic prediction confidence interval are used as inputs to the net load prediction interval. The core processing of this invention is not limited to a specific prediction model. As long as the predicted load value of the power supply zone, the predicted photovoltaic output value, the load prediction confidence interval, and the photovoltaic prediction confidence interval can be obtained, the net load prediction interval generation, key scheduling window identification, energy storage state of charge dynamic adjustment corridor generation, energy storage charge and discharge plan generation, rolling correction, and execution described in Embodiment 1 can be performed.

[0029] Example 3: Based on Example 1, this example describes the upper-level scheduling and coordination method when the energy storage regulation capacity is insufficient. When the energy management platform determines in S3 that the lower limit of the dynamic adjustment corridor for the state of charge of energy storage is higher than the upper limit, it means that the energy storage cannot simultaneously meet the charging space required by the subsequent photovoltaic consumption window and the discharge capacity required by the demand peak shaving window within the current scheduling cycle. The energy management platform prioritizes the discharge capacity required by the demand peak shaving window and generates a peak shaving priority identifier. For the surplus photovoltaic power that cannot be absorbed by energy storage due to peak shaving priority, the energy management platform generates a quantity to be coordinated. The quantity to be coordinated includes the time period to be coordinated, the amount of electricity to be coordinated, the corresponding power supply zone, and the cause identifier; the cause identifier is used to distinguish between insufficient energy storage capacity, insufficient energy storage charging power, energy storage state of charge reaching the upper limit, unavailable energy storage equipment, or communication abnormality. The amount of electricity to be coordinated is determined by the difference between the surplus photovoltaic power to be consumed within the photovoltaic consumption window and the energy storage available charging power. Specifically, within the photovoltaic consumption window, the surplus photovoltaic power to be consumed is determined based on the surplus power corresponding to the net load forecast upper limit being less than zero and the scheduling time step. The energy storage available charging power is determined based on the actual charging space of the energy storage, the maximum charging power of the energy storage, the charging efficiency, and the scheduling time step. When the surplus photovoltaic power to be consumed is greater than the energy storage available charging power, the difference between the two is output to the upper-level scheduling system as the amount of electricity to be coordinated. The peak shaving power gap is determined by the difference between the power demand exceeding the demand control target within the demand shaving window and the energy storage's executable discharge power. Specifically, within the demand shaving window, the power demand exceeding the demand control target is determined based on the upper limit of the net load forecast or the difference between the actual net load and the demand control target. The energy storage's executable discharge power is determined based on the actual dischargeable energy storage capacity, the maximum discharge power of the energy storage, the discharge efficiency, and the lower limit of the energy storage's state of charge dynamic adjustment corridor. When the power demand exceeding the demand control target is greater than the energy storage's executable discharge power, the difference between the two is output to the upper-level dispatch system as the peak shaving power gap. After receiving the quantity to be coordinated, the upper-level dispatch system queries the orderly charging capacity of charging piles, the flexible load capacity of air conditioning, and the cross-regional absorption capacity of adjacent power supply zones according to the resource call order. If there are charging piles with orderly charging capacity in the power supply zone, the upper-level dispatch system increases the charging power of the charging piles or arranges charging in advance without affecting the vehicle's usage demand. If it is a cooling or heating scenario and the flexible load of air conditioning has adjustment conditions, the upper-level dispatch system calls on air conditioning pre-cooling, heat pumps to be turned on, or other flexible loads to absorb some of the surplus electricity. If the resources in this power supply zone are insufficient and the adjacent power supply zones have absorption capacity, the upper-level dispatch system transfers some of the surplus electricity through cross-regional mutual assistance. When the energy management platform encounters a peak-shaving power shortfall in S4 or S6, the upper-level dispatch system processes it according to the principle of prioritizing demand control targets. If the energy storage discharge power is insufficient to control the grid power intake within the demand control target, the upper-level dispatch system calls on flexible air conditioning adjustment, charging pile power reduction, or other interruptible loads to reduce the power intake of the power supply zone. Through this collaborative approach, a complete dispatch is still formed even when energy storage cannot independently complete the dispatch target.

[0030] Example 4: Based on Example 1, this example explains the execution relationship between intraday rolling correction and real-time correction. During the day-ahead phase, the energy management platform executes S1 to S4 based on the power supply zone load forecast, photovoltaic output forecast, load forecast confidence interval, photovoltaic forecast confidence interval, and energy storage operation status within the future scheduling cycle to generate the day-ahead energy storage charging and discharging plan. The day-ahead energy storage charging and discharging plan is used to determine the energy storage action type and power target for each time period within the future scheduling cycle. During the intraday rolling phase, the energy management platform obtains the latest actual net load, the latest forecast data, and the actual state of charge of energy storage according to the rolling correction cycle. If the actual net load deviates continuously from the net load forecast range within the sampling sequence corresponding to the rolling correction cycle, the energy management platform re-executes S1 to S4 to generate a corrected energy storage charging and discharging plan using the actual state of charge of energy storage as the new initial state. The corrected energy storage charging and discharging plan replaces the unexecuted portion of the original plan. During the real-time correction phase, the energy management platform or energy storage management system fine-tunes the current energy storage power command based on real-time measurement data. Real-time correction does not change the planned structure for all future time periods and is mainly used to smooth out short-term fluctuations in photovoltaic power and sudden load changes. The current energy storage power after real-time correction still meets the restrictions of the maximum charging power of energy storage, the maximum discharging power of energy storage, the allowable state of charge range of energy storage, and the dynamic adjustment corridor of the state of charge of energy storage. By employing a layered execution approach of daily planning, intraday rolling, and real-time correction, this embodiment can not only utilize daily forecast results to form a complete energy storage dispatch plan, but also promptly correct deviations in photovoltaic output, passenger load, or air conditioning load, thus avoiding the continuous accumulation of forecast errors that could lead to mismatch in energy storage regulation capacity.

[0031] This invention also proposes a photovoltaic-storage coordinated energy dispatching system based on net load prediction, used to execute a photovoltaic-storage coordinated energy dispatching method based on net load prediction as described in Embodiment 1, comprising: The prediction interval generation module is used to acquire power supply zone prediction data and energy storage operation status, and generate net load prediction intervals based on the power supply zone prediction data. The critical window identification module is used to identify critical scheduling windows based on the net load forecast interval. The critical scheduling windows include photovoltaic consumption windows, demand shaving windows, and uncertain reversal windows. The dynamic corridor generation module is used to generate a dynamic adjustment corridor for energy storage state of charge (ESC) to limit the upper and lower limits of ESC operation based on the photovoltaic absorption window and the demand peak shaving window. The charge / discharge plan generation module is used to generate an energy storage charge / discharge plan based on the energy storage state of charge, the dynamic adjustment corridor, the net load prediction range, and the energy storage operation status. The rolling correction module is used to obtain the actual net load deviation and to perform rolling correction on the energy storage charging and discharging plan based on the actual net load deviation to obtain the corrected energy storage charging and discharging plan. The instruction module is used to convert the energy storage charge and discharge plan or the modified energy storage charge and discharge plan into energy storage power instructions.

[0032] The above description is only a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any equivalent substitutions or modifications made by those skilled in the art within the scope of the technology disclosed in the present invention, based on the technical solution and inventive concept of the present invention, should be covered within the scope of protection of the present invention.

Claims

1. A photovoltaic-storage coordinated energy dispatch method based on net load forecasting, characterized in that, Includes the following steps: S1. Obtain power supply zone prediction data and energy storage operation status, and generate net load prediction intervals based on the power supply zone prediction data; S2. Identify key scheduling windows based on the net load forecast interval. The key scheduling windows include photovoltaic consumption window, demand shaving window and uncertain reversal window. S3. Generate a dynamic adjustment corridor for the energy storage state of charge (ESC) to limit the upper and lower limits of the ESC operation based on the photovoltaic absorption window and the demand peak shaving window. S4. Generate an energy storage charging and discharging plan based on the energy storage state of charge, the dynamic adjustment corridor, the net load prediction range, and the energy storage operation status. S5. Obtain the actual net load deviation, and perform rolling correction on the energy storage charging and discharging plan based on the actual net load deviation to obtain the corrected energy storage charging and discharging plan. S6. Convert the energy storage charging and discharging plan or the modified energy storage charging and discharging plan into an energy storage power command.

2. The photovoltaic-storage coordinated energy dispatch method based on net load forecasting according to claim 1, characterized in that, The power supply zone prediction data includes the power supply zone load prediction value, photovoltaic output prediction value, load prediction confidence interval, photovoltaic prediction confidence interval, time-of-use electricity price information, and demand control target; the energy storage operation status includes the energy storage state of charge, energy storage rated capacity, energy storage maximum charging power, energy storage maximum discharging power, charging efficiency, discharging efficiency, and energy storage allowable state of charge range.

3. The photovoltaic-storage coordinated energy dispatch method based on net load forecasting according to claim 2, characterized in that, The net load forecast interval is generated based on the power supply zone forecast data, including: The net load forecast value is generated based on the difference between the predicted load value of the power supply zone and the predicted photovoltaic output value, and the net load forecast interval is generated based on the load forecast confidence interval and the photovoltaic forecast confidence interval. The net load forecast interval includes a lower bound and an upper bound. The lower bound is determined by the lower bound of the load forecast and the upper bound of the photovoltaic forecast, and the upper bound is determined by the upper bound of the load forecast and the lower bound of the photovoltaic forecast.

4. The photovoltaic-storage coordinated energy dispatch method based on net load forecasting according to claim 3, characterized in that, Identifying key scheduling windows based on the net load forecast interval includes: When the upper limit of the net load forecast is less than zero, the corresponding time period will be marked as the photovoltaic consumption window; When the upper limit of the net load forecast is greater than the demand control target, the corresponding time period will be marked as the demand shaving window. When the lower bound of net load forecast is less than zero and the upper bound of net load forecast is greater than zero, the corresponding time period is marked as an uncertain inversion window.

5. The photovoltaic-storage coordinated energy dispatch method based on net load forecasting according to claim 4, characterized in that, Based on the photovoltaic absorption window and the demand peak shaving window, a dynamic adjustment corridor for the energy storage state of charge (SBC) is generated to limit the upper and lower limits of the SBC operation, including: The amount of rechargeable space that should be reserved by energy storage before entering the photovoltaic absorption window is determined based on the surplus photovoltaic power within the photovoltaic absorption window. The amount of dischargeable energy that the energy storage should retain before entering the demand shaving window is determined based on the power demand exceeding the demand control target within the demand shaving window. A dynamic adjustment corridor for the energy storage state of charge is generated based on the rechargeable space, the dischargeable capacity, and the allowable state of charge range.

6. The photovoltaic-storage coordinated energy dispatch method based on net load forecasting according to claim 5, characterized in that, When the lower limit of the energy storage state of charge dynamically adjusts the corridor based on the rechargeable space and the dischargeable power, which is higher than the upper limit, the dischargeable power required by the demand peak shaving window is retained, and the unmet photovoltaic power consumption is output to the upper-level scheduling system as a quantity to be coordinated.

7. The photovoltaic-storage coordinated energy dispatch method based on net load forecasting according to claim 5, characterized in that, Based on the energy storage state of charge dynamic adjustment corridor, net load prediction range, and energy storage operating status, an energy storage charging and discharging plan is generated, including: When the current time period falls within the photovoltaic consumption window and the energy storage state of charge is below the upper limit of the energy storage state of charge dynamic adjustment corridor, an energy storage charging plan is generated. When the current time period falls within the demand shaving window and the energy storage state of charge is higher than the lower limit of the energy storage state of charge dynamic adjustment corridor, an energy storage discharge plan is generated. When the current time period is neither a photovoltaic consumption window nor a demand peak shaving window, a charging plan or discharging plan is generated to bring the energy storage state of charge back into the energy storage state of charge dynamic adjustment corridor, or a standby plan is generated to keep the energy storage in the energy storage state of charge dynamic adjustment corridor, based on the position of the energy storage state of charge relative to the energy storage state of charge dynamic adjustment corridor.

8. A photovoltaic-storage coordinated energy dispatching method based on net load forecasting according to claim 7, characterized in that, Obtain the actual net load deviation, and perform rolling adjustments to the energy storage charging and discharging plan based on the actual net load deviation, including: The actual load, actual photovoltaic output, and actual state of charge of energy storage in the power supply zone are obtained, and the actual net load is generated based on the difference between the actual load and the actual photovoltaic output of the power supply zone. The actual net load deviation is generated based on the difference between the actual net load and the predicted net load. When the actual net load deviates continuously from the net load prediction range within the sampling sequence corresponding to the rolling correction cycle, or when the actual net load within the uncertain reversal window has been determined to be in a positive or negative state, the energy storage charging and discharging plan for the subsequent time period is regenerated based on the latest actual net load and the actual state of charge of the energy storage.

9. A photovoltaic-storage coordinated energy dispatch method based on net load forecasting according to claim 8, characterized in that, Converting the energy storage charge / discharge plan or the modified energy storage charge / discharge plan into energy storage power commands includes: The energy storage charging and discharging plan or the revised energy storage charging and discharging plan will be sent to the energy storage energy management system or the energy storage converter. Obtain the actual charging and discharging power, energy storage state of charge, and equipment operating status fed back by the energy storage management system or energy storage converter; When the feedback result indicates that the energy storage cannot execute the energy storage power command, the executable energy storage power command for the current time period is re-determined based on the actual available energy storage power, and the unmet photovoltaic power consumption or peak shaving power gap is output to the upper-level dispatch system.

10. A photovoltaic-storage coordinated energy dispatching system based on net load forecasting, applied to the photovoltaic-storage coordinated energy dispatching method based on net load forecasting as described in any one of claims 1-9, characterized in that, include: The prediction interval generation module is used to acquire power supply zone prediction data and energy storage operation status, and generate net load prediction intervals based on the power supply zone prediction data. The critical window identification module is used to identify critical scheduling windows based on the net load forecast interval. The critical scheduling windows include photovoltaic consumption windows, demand shaving windows, and uncertain reversal windows. The dynamic corridor generation module is used to generate a dynamic adjustment corridor for energy storage state of charge (ESC) to limit the upper and lower limits of ESC operation based on the photovoltaic absorption window and the demand peak shaving window. The charge / discharge plan generation module is used to generate an energy storage charge / discharge plan based on the energy storage state of charge, the dynamic adjustment corridor, the net load prediction range, and the energy storage operation status. The rolling correction module is used to obtain the actual net load deviation and to perform rolling correction on the energy storage charging and discharging plan based on the actual net load deviation to obtain the corrected energy storage charging and discharging plan. The instruction module is used to convert the energy storage charge and discharge plan or the modified energy storage charge and discharge plan into energy storage power instructions.

Images