Energy storage charging and discharging monitoring and auxiliary decision-making method and system under multiple time scales

Abstract

The invention discloses an energy storage charging and discharging monitoring and auxiliary decision-making method and system under multiple time scales, and the method comprises the steps: carrying out the collection and verification of the operation state of an energy storage station group, and forming a chargeable / dischargeable capability surface based on the province-based and whole-network aggregation; a suggested target curve is generated on multiple time scales; when it is predicted that new energy peak regulation blocking occurs in the noon, the system calculates the needed charging power; when a peak segment balance gap is predicted and displayed, the system calculates a discharge strategy; after the strategy is generated, checking the cross-section tidal current and the climbing rate, and reserving the peak electric quantity of the next day in a cross-day scene; and the suggested curve lands according to the hierarchy of whole network-provincial-zoning-station. According to the method, energy storage resources can be arranged in a unified manner in multiple time scales, multi-level landing can be oriented in a master station platform, online rolling check is supported, and new energy consumption, peak section guarantee supply and section safety are considered at the same time.

Description

Technical Field

[0001] This invention belongs to the field of power system dispatching and control technology, and particularly relates to a method and system for monitoring and assisting decision-making of energy storage charging and discharging under multiple time scales. Background Technology

[0002] With a high proportion of renewable energy connected to the grid, the system simultaneously faces fluctuations in renewable energy output and uncertainties on the load side, leading to tight conventional peak-shaving resources and significant resource mismatch across different sections. Traditional strategies based on manual experience or a single time scale are insufficient to uniformly consider renewable energy consumption, peak-shaving supply, and section safety in cross-day / day-ahead / intra-day scenarios. Furthermore, the allocation of power station clusters lacks verifiable overall rules, making execution and backtracking difficult.

[0003] To address the power output fluctuations and load uncertainties resulting from the high proportion of renewable energy integration, existing solutions have attempted to employ "day-ahead-intraday" multi-timescale collaborative optimization, coordinating energy storage with conventional power sources and demand-side resources within a unified model. For example, the literature ("Day-ahead-intraday Collaborative Optimization Scheduling Strategy for Photovoltaic-Storage Joint Systems Considering Peak Electricity Price Compensation Mechanisms," Yin Xiaodong et al., Journal of Energy and Power Engineering, 2024) proposes a "day-ahead-intraday joint optimization" strategy for photovoltaic-storage power plants. By coupling two timescales within the same optimization framework, it mitigates the interference of day-ahead forecast deviations on operation. Patent document CN117895497B discloses a "Multi-Timescale Collaborative Optimization Scheduling Method and System," which simultaneously considers mobile energy storage and power flow constraints on the distribution side, unifying day-ahead and intraday rescheduling into a unified model to improve economic efficiency and absorption capacity. These solutions demonstrate that layered timescale energy storage scheduling helps reduce wind and solar curtailment, mitigate fluctuations, and reduce network losses.

[0004] However, the existing technologies mentioned above are mostly focused on joint optimization of parks / distribution networks or single regions, and still have the following problems: (1) There is a lack of verifiable allocation rules that are consistent across multiple levels of "whole network - province - zone - station", and the charging and discharging order, cut-off and rollback rules between station groups are difficult to audit; (2) Safety constraints such as cross-section (connection line / restricted channel) power flow and ramp rate are mostly reflected by offline verification or experience thresholds, which are difficult to support online rolling closed loop; (3) The disclosure of "next day peak" retention management and cross-regional mutual assistance strategies in cross-day scenarios is insufficient, and the regional boundary conditions for "on-site charging / peak discharge" are not agreed upon in a consistent manner; (4) The source of predicted values ​​and the online correction mechanism are not deeply coupled with the main station platform, resulting in the lag in strategy issuance under the "offline calculation + manual interpretation" form.

[0005] Therefore, there is still a need for a method and system that can uniformly orchestrate energy storage resources across multiple time scales, be implemented at multiple levels on the main station platform, and support online rolling verification, so as to simultaneously take into account renewable energy consumption, peak supply guarantee, and cross-sectional safety. Summary of the Invention

[0006] To address the shortcomings of existing technologies, this invention provides a method and system for monitoring and assisting decision-making for energy storage charging and discharging across multiple time scales. The aim is to construct a "multi-time scale—full-level—rolling optimization" energy storage charging and discharging monitoring and assisting decision-making system: unified generation of charging / discharging targets under intraday, day-ahead, and 3–5 day rolling outlooks; realization of cross-level coordination and refined allocation of station clusters under network-wide constraints and section safety conditions; and integration with SCADA / WAMS / planning databases within the D5000 and other scheduling master station platform environments, forming an engineering solution capable of online closed-loop verification and traceability.

[0007] The present invention adopts the following technical solution.

[0008] This invention proposes a method for monitoring and assisting decision-making in energy storage charging and discharging across multiple time scales, including: Collect operational status data from each energy storage station, calculate the station-level chargeable / dischargeable capacity based on the operational status data of each energy storage station, and summarize the calculated station-level chargeable / dischargeable capacity to the entire network to obtain the chargeable / dischargeable capacity surface; Obtain station-level variables, including charging / discharging power and energy; based on the chargeable / discharging capacity and the charging / discharging power and energy, set constraints, including energy balance, power / energy boundary and ramp constraints, mutual exclusion logic constraints, remaining curtailment constraints, and remaining supply-demand balance gap constraints; design an objective function based on the trade-off between curtailment / gap / cycle cost. Based on the objective function and constraints, the following are obtained: the whole-grid target curve and charging suggestion curve considering the obstruction of new energy peak shaving at multiple time scales, and the whole-grid target curve and discharging suggestion curve considering the grid balance gap. The cross-sectional power flow and ramp rate are checked on the charging and discharging recommendation curves. The power flow changes caused by the energy storage charging and discharging behavior are simulated. If voltage exceeds the limit, branch overload or cross-sectional overload occurs, the whole-network target curve and charging / discharging recommendation curve for the new energy peak shaving obstruction and grid balance gap are adjusted according to the actual situation until the constraints are met. After the verification is completed, the peak power of the next day is reserved in the cross-day scenario. Based on the grid-wide target curve of the obstruction of new energy peak shaving and the grid balance gap, and the charging / discharging strategy of reserving the peak power for the next day, the charging and discharging strategy is allocated and executed in the order of grid-province-region-station.

[0009] More preferably, the mutual exclusion logic is implemented using a large M + 0 / 1 variable to constrain the charging / discharging power; The objective function based on the trade-off between curtailment / shortage / circular cost is in the full time domain. Minimize the curtailment surplus Curt, the supply-demand balance gap surplus Gap+, and the cycle equivalent cost. The weighted sum; The remaining amount of abandoned electricity Curt, the remaining amount of supply and demand balance gap Gap+, are related to the predicted value of renewable energy power, the minimum total output that conventional units can reduce to, the maximum total output that conventional units can increase, the predicted value of load power, the power transmitted to other units, and the charging / discharging power of all stations.

[0010] The weights are given by the operating strategy preferences.

[0011] More preferably, the specific steps for obtaining the grid-wide target curves for two control scenarios—namely, the obstruction of renewable energy peak shaving and the grid balance gap—based on the objective function and constraints, are as follows: For day-to-day / multi-day scenarios, mixed integer linear programming (MILP) is used to solve the problem and obtain a globally feasible baseline charge / discharge target. For intraday rolling scenarios, while maintaining the consistency of constraints, we can inherit MILP and resolve it with a rolling window, or replace the mutual exclusion logic with soft constraints / large penalty coefficients, degenerating into linear programming LP or convex programming for fast recalculation. The solution yields the network-wide target curves for the obstruction of new energy peak shaving and the grid balance gap.

[0012] More preferably, the network-wide target curve for the theoretical peak-shaving of new energy sources being hindered is represented as follows: The sum of the predicted value of renewable energy power and the minimum total output that can be reduced to conventional units is calculated. The sum is then subtracted from the predicted value of system load power and the power transmitted to the outside to obtain the difference result. The larger of 0 and the difference result is used to characterize the theoretical peak-shaving obstruction curve of renewable energy. The target curve for the entire power grid balance gap is represented as follows: The sum of the predicted system load power and the transmitted power is calculated. The sum is then subtracted from the predicted new energy power and the maximum total output that can be raised by conventional units to obtain the difference result. The larger of 0 and the difference result is used to characterize the grid-wide target curve of the grid balance gap.

[0013] More preferably, the method for obtaining the predicted value of the new energy power is as follows: The hourly forecasts are generated using the baseline method of numerical weather prediction (NWP) and power curves of power plants. Within the day, the deviation between measured power output and meteorological factors is corrected by Kalman / exponential weighting using a near-rolling window to ensure rolling updates. The predicted new energy power values ​​are time-aligned, missing data are imputed, and out-of-bounds data are removed before being entered into the database. The method for obtaining the predicted load power value is as follows: A regression / tree model was recently established using historical load, meteorological and calendar factors to form an hourly baseline, which is then rolled over within a set time window based on the latest measured deviation.

[0014] More preferably, based on the network-wide target curve indicating that peak shaving is hindered by the new energy theory, a recommended charging power is generated: The recommended charging power is subject to three constraints, including the upper limit of the blocked power (Cte); and the upper limit of the device's charging power. The upper bound of the power that the upper limit energy space can absorb within the current step size Δt; the minimum value among the three constraints is used as the suggested charging power; Based on the grid-wide target curve for the aforementioned grid balance gap, a recommended discharge power is generated: It is recommended that the discharge power also be subject to triple constraints, including the upper limit of the gap power; and the upper limit of the device discharge power. The upper bound of the power that can be released within the current step size in the lower limit energy space; the minimum value among the three constraints is used as the suggested discharge power.

[0015] More preferably, the method for reserving peak electricity for the next day in a cross-day scenario is as follows: Recently—For several days: The rechargeable / dischargeable capacity is used as the adjustable capacity. Based on the prediction and adjustable capacity, the baseline curve and the reserved peak power for the next day are obtained. The forecasts include day-ahead / multi-day baseline forecasts and intraday rolling corrections of system load power forecasts and renewable energy power forecasts, as well as the loading of external power and the lower limit of conventional unit capacity. The adjustable capabilities also include cross-sectional power flow limit, minimum total output that can be reduced to by conventional units, and maximum total output that can be raised by conventional units; Based on the predictive and adjustable capabilities, MILP is used for day-ahead / multi-day charging and discharging, and MILP rolling or degraded LP is used for intraday charging and discharging. The model is solved in the same time domain as the predictive and adjustable capabilities to obtain the baseline charging / discharging curve and the reserved peak power for the next day in cross-day scenarios. The baseline charge / discharge curve refers to the hourly charge / discharge power reference curve used to guide intraday rolling optimization within the daytime / multi-day time domain, including the time period and power allocation information for reserving peak power for the next day in cross-day scenarios; Intraday: Update forecasts and status using a rolling window, resolvable or quickly redistributed.

[0016] More preferably, the charging and discharging strategy is allocated and executed in the order of network-province-region-station: Network-wide: Generate the overall target charge / discharge curve for the entire network: Following the principle of global optimal scheduling, the entire network will not discharge when renewable energy is blocked; charging will not be carried out when the entire network is short-circuited; charging will not be arranged outside the section until thermal power reaches its minimum technical output; and local charging within the section is allowed when the section is heavily loaded, provided that it does not affect the support of the entire network. By province: The overall network target is broken down into provincial targets based on the proportion of energy storage capacity in each province and the constraints of provincial sections; Zoning: The provincial objectives are refined into zonal objectives based on zonal constraints and local capabilities; Station-level: Precisely assigning zone targets to individual stations: Charging allocation: Stations are sorted by SOC% from smallest to largest, with priority given to fully charging stations with low SOC, while adhering to regional targets and station limits; Discharge allocation: Sorted by station SOC% from largest to smallest, with priority given to stations with high SOC for peak discharge; Independent energy storage takes precedence over integrated energy storage; when there is insufficient or overloaded capacity, it should be cut off according to capacity and constraints.

[0017] This invention also proposes a multi-timescale energy storage charging and discharging monitoring and auxiliary decision-making system utilizing the above method, comprising a data and model layer, a business layer, and a presentation layer: The data and model layer loads station cluster operation data as the raw sequence; timestamp alignment, missing data imputation, and out-of-bounds removal are performed on the raw sequence to obtain cleaned time-series data; and the inflatable / expandable capacity surface is calculated. At the business layer, station-level variables are acquired, including charging / discharging power and energy. Constraints are set based on the available charging / discharging capacity and the charging / discharging power and energy. An objective function is designed based on a trade-off between power curtailment, power shortage, and cycle cost. Based on the objective function and constraints, a network-wide target curve considering multiple time scales of renewable energy peak-shaving obstruction and grid balance gaps is obtained, along with suggested charging and discharging curves. The suggested curves are verified for cross-sectional power flow and power ramp-up rate. In cross-day scenarios, power reserve and allocation are performed based on the next day's peak demand and the current day's charging / discharging schedule. Paths are allocated in the order of network-province-region-station, following SOC sorting and station-level power / energy / mutual exclusion / ramp-up constraints, refining regional targets into station-level target power. The display layer provides a multi-dimensional perspective of panoramic / province / region / station, showing the real-time status and strategy execution, and statistically analyzing peak shaving power, curtailment rate, supply-demand gap reduction, charge-discharge ramp-up, and cycle cost indicators; it supports historical traceability and offline scoring; and it performs offline evaluation of strategy feasibility and performance and stores the results in the database for parameter solidification and reporting.

[0018] More preferably, the system runs on a Linux+D5000 environment on the main station side, with the backend C / C++ implementing the core computing and the frontend Java; the external interfaces include SCADA / WAMS, planning and forecasting libraries, section / connection line quota and EMS relationship libraries; and a layered writing library for policy, execution and verification logs to support online closed loop and traceability.

[0019] Compared with the prior art, the beneficial effects of the present invention are as follows: 1. This invention integrates multiple time scales and for the first time generates energy storage charging / discharging targets within a unified framework of "day-to-multi-day-intra-day rolling," while also considering the reserved power capacity for the "next day peak" and the cross-day allocation ratio; 2. This invention integrates cross-section safety constraints and network-wide strategy rules through global optimization and cross-section constraint coordination, taking into account both renewable energy consumption and peak supply guarantee. 3. This invention enables refined allocation of station groups: based on SOC percentage and station-level capacity / locking signals, an executable sequence of "priority-truncation-back" is formed; 4. This invention is verifiable and feasible, and by providing database structure, interface and function implementation references, it supports the online commissioning of D5000; 5. This invention can be robustly extended: by introducing scenario sets or opportunity constraints to model source load uncertainty, and balancing the risk of power curtailment / shortage with the cycle cost. Attached Figure Description

[0020] Figure 1 This is a flowchart of the method for monitoring and assisting decision-making of energy storage charging and discharging under multiple time scales according to the present invention; Figure 2 This invention generates a business process diagram using a multi-scale strategy. Figure 3 This is a curve showing the obstruction of peak shaving in the new energy theory of this invention; Figure 4 This is a suggested energy storage charging curve for the present invention; Figure 5 This is the spare curve / balance gap diagram of the present invention; Figure 6 This is a suggested energy storage discharge curve for the present invention; Figure 7 This is a structural diagram of the strategy target allocation implementation of the present invention; Figure 8 This is a schematic diagram of the overall system architecture of the present invention. Detailed Implementation

[0021] To make the objectives, technical solutions, and advantages of this invention clearer, the technical solutions of this invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of this invention. The embodiments described in this application are merely some embodiments of this invention, and not all embodiments. Based on the spirit of this invention, all other embodiments obtained by those skilled in the art without creative effort are within the protection scope of this invention.

[0022] The present invention proposes the following technical solution: This invention proposes a method for monitoring and assisting decision-making in energy storage charging and discharging across multiple time scales, including: Collect operational status data from each energy storage station, calculate the station-level chargeable / dischargeable capacity based on the operational status data of each energy storage station, and summarize the calculated station-level chargeable / dischargeable capacity to the entire network to obtain the chargeable / dischargeable capacity surface; Obtain station-level variables, including charging / discharging power and energy; based on the chargeable / discharging capacity and the charging / discharging power and energy, set constraints, including energy balance, power / energy boundary and ramp constraints, mutual exclusion logic constraints, remaining curtailment constraints, and remaining supply-demand balance gap constraints; design an objective function based on the trade-off between curtailment / gap / cycle cost. The mutual exclusion logic is implemented using a large M + 0 / 1 variable to constrain the charging / discharging power; The objective function based on the trade-off between curtailment / shortage / circular cost is in the full time domain. Minimize the curtailment surplus Curt, the supply-demand balance gap surplus Gap+, and the cycle equivalent cost. The weighted sum; The remaining amount of abandoned electricity Curt, the remaining amount of supply and demand balance gap Gap+, are related to the predicted value of renewable energy power, the minimum total output that conventional units can reduce to, the maximum total output that conventional units can increase, the predicted value of load power, the power transmitted to other units, and the charging / discharging power of all stations.

[0023] The weights are given by the operating strategy preferences.

[0024] Based on the objective function and constraints, the following are obtained: the whole-grid target curve and charging suggestion curve considering the obstruction of new energy peak shaving at multiple time scales, and the whole-grid target curve and discharging suggestion curve considering the grid balance gap. Based on the objective function and constraints, the network-wide target curves for two control scenarios—namely, the obstruction of renewable energy peak shaving and the grid balance gap—are obtained. The specific steps are as follows: For day-to-day / multi-day scenarios, mixed integer linear programming (MILP) is used to solve the problem and obtain a globally feasible baseline charge / discharge target. For intraday rolling scenarios, while maintaining the consistency of constraints, we can inherit MILP and resolve it with a rolling window, or replace the mutual exclusion logic with soft constraints / large penalty coefficients, degenerating into linear programming LP or convex programming for fast recalculation. The solution yields the network-wide target curves for the obstruction of new energy peak shaving and the grid balance gap.

[0025] The network-wide target curve for the obstruction of peak shaving in the new energy theory is represented as follows: The sum of the predicted value of renewable energy power and the minimum total output that can be reduced to conventional units is calculated. The sum is then subtracted from the predicted value of system load power and the power transmitted to the outside to obtain the difference result. The larger of 0 and the difference result is used to characterize the theoretical peak-shaving obstruction curve of renewable energy. The target curve for the entire power grid balance gap is represented as follows: The sum of the predicted system load power and the transmitted power is calculated. The sum is then subtracted from the predicted new energy power and the maximum total output that can be raised by conventional units to obtain the difference result. The larger of 0 and the difference result is used to characterize the grid-wide target curve of the grid balance gap.

[0026] The method for obtaining the predicted value of the new energy power is as follows: The hourly forecasts are generated using the baseline method of numerical weather prediction (NWP) and power curves of power plants. Within the day, the deviation between measured power output and meteorological factors is corrected by Kalman / exponential weighting using a near-rolling window to ensure rolling updates. The predicted new energy power values ​​are time-aligned, missing data are imputed, and out-of-bounds data are removed before being entered into the database. The method for obtaining the predicted load power value is as follows: A regression / tree model was recently established using historical load, meteorological and calendar factors to form an hourly baseline, which is then rolled over within a set time window based on the latest measured deviation.

[0027] Based on the network-wide target curve indicating that peak shaving is hindered by the aforementioned new energy theory, the following recommended charging power is generated: The recommended charging power is subject to three constraints, including the upper limit of the blocked power (Cte); and the upper limit of the device's charging power. The upper bound of the power that the upper limit energy space can absorb within the current step size Δt; the minimum value among the three constraints is used as the suggested charging power; Based on the grid-wide target curve for the aforementioned grid balance gap, a recommended discharge power is generated: It is recommended that the discharge power also be subject to triple constraints, including the upper limit of the gap power; and the upper limit of the device discharge power. The upper bound of the power that can be released within the current step size in the lower limit energy space; the minimum value among the three constraints is used as the suggested discharge power.

[0028] The cross-sectional power flow and ramp rate are checked on the charging and discharging recommendation curves. The power flow changes caused by the energy storage charging and discharging behavior are simulated. If voltage exceeds the limit, branch overload or cross-sectional overload occurs, the whole-network target curve and charging / discharging recommendation curve for the new energy peak shaving obstruction and grid balance gap are adjusted according to the actual situation until the constraints are met. After the verification is completed, the peak power of the next day is reserved in the cross-day scenario. The specific method for reserving peak electricity for the next day in cross-day scenarios is as follows: Recently—For several days: The rechargeable / dischargeable capacity is used as the adjustable capacity. Based on the prediction and adjustable capacity, the baseline curve and the reserved peak power for the next day are obtained. The forecasts include day-ahead / multi-day baseline forecasts and intraday rolling corrections of system load power forecasts and renewable energy power forecasts, as well as the loading of external power and the lower limit of conventional unit capacity. The adjustable capabilities also include cross-sectional power flow limit, minimum total output that can be reduced to by conventional units, and maximum total output that can be raised by conventional units; Based on the predictive and adjustable capabilities, MILP is used for day-ahead / multi-day charging and discharging, and MILP rolling or degraded LP is used for intraday charging and discharging. The model is solved in the same time domain as the predictive and adjustable capabilities to obtain the baseline charging / discharging curve and the reserved peak power for the next day in cross-day scenarios. The baseline charge / discharge curve refers to the hourly charge / discharge power reference curve used to guide intraday rolling optimization within the daytime / multi-day time domain, including the time period and power allocation information for reserving peak power for the next day in cross-day scenarios; Intraday: Update forecasts and status using a rolling window, resolvable or quickly redistributed.

[0029] Based on the grid-wide target curve of the obstruction of new energy peak shaving and the grid balance gap, and the charging / discharging strategy of reserving the peak power for the next day, the charging and discharging strategy is allocated and executed in the order of grid-province-region-station.

[0030] The charging and discharging strategy is allocated and executed in the order of network-wide, province-by-region, and station-by-station. Network-wide: Generate the overall target charge / discharge curve for the entire network: Following the principle of global optimal scheduling, the entire network will not discharge when renewable energy is blocked; charging will not be carried out when the entire network is short-circuited; charging will not be arranged outside the section until thermal power reaches its minimum technical output; and local charging within the section is allowed when the section is heavily loaded, provided that it does not affect the support of the entire network. By province: The overall network target is broken down into provincial targets based on the proportion of energy storage capacity in each province and the constraints of provincial sections; Zoning: The provincial objectives are refined into zonal objectives based on zonal constraints and local capabilities; Station-level: Precisely assigning zone targets to individual stations: Charging allocation: Stations are sorted by SOC% from smallest to largest, with priority given to fully charging stations with low SOC, while adhering to regional targets and station limits; Discharge allocation: Sorted by station SOC% from largest to smallest, with priority given to stations with high SOC for peak discharge; Independent energy storage takes precedence over integrated energy storage; when there is insufficient or overloaded capacity, it should be cut off according to capacity and constraints.

[0031] This invention also proposes a multi-timescale energy storage charging and discharging monitoring and auxiliary decision-making system utilizing the above method, comprising a data and model layer, a business layer, and a presentation layer: The data and model layer loads station cluster operation data as the raw sequence; timestamp alignment, missing data imputation, and out-of-bounds removal are performed on the raw sequence to obtain cleaned time-series data; and the inflatable / expandable capacity surface is calculated. At the business layer, station-level variables are acquired, including charging / discharging power and energy. Constraints are set based on the available charging / discharging capacity and the charging / discharging power and energy. An objective function is designed based on a trade-off between power curtailment, power shortage, and cycle cost. Based on the objective function and constraints, a network-wide target curve considering multiple time scales of renewable energy peak-shaving obstruction and grid balance gaps is obtained, along with suggested charging and discharging curves. The suggested curves are verified for cross-sectional power flow and power ramp-up rate. In cross-day scenarios, power reserve and allocation are performed based on the next day's peak demand and the current day's charging / discharging schedule. Paths are allocated in the order of network-province-region-station, following SOC sorting and station-level power / energy / mutual exclusion / ramp-up constraints, refining regional targets into station-level target power. The display layer provides a multi-dimensional perspective of panoramic / province / region / station, showing the real-time status and strategy execution, and statistically analyzing peak shaving power, curtailment rate, supply-demand gap reduction, charge-discharge ramp-up, and cycle cost indicators; it supports historical traceability and offline scoring; and it performs offline evaluation of strategy feasibility and performance and stores the results in the database for parameter solidification and reporting.

[0032] More preferably, the system runs on a Linux+D5000 environment on the main station side, with the backend C / C++ implementing the core computing and the frontend Java; the external interfaces include SCADA / WAMS, planning and forecasting libraries, section / connection line quota and EMS relationship libraries; and a layered writing library for policy, execution and verification logs to support online closed loop and traceability.

[0033] Example 1 This invention proposes a method for monitoring and assisting decision-making regarding energy storage charging and discharging across multiple time scales, such as... Figure 1As shown, the system first collects and verifies the operational status of the energy storage station cluster, including SOC, capacity upper and lower limits, charging and discharging power limits, efficiency parameters, and charging / discharging lockout status. Based on this, a "chargeable / discharging capacity surface" is formed, aggregated by province and the entire network. Then, suggested target curves are generated at multiple time scales: when forecasts indicate that peak shaving by new energy sources is hindered at midday, the system calculates the required charging power and energy to suppress power curtailment; when forecasts indicate a peak balance gap, the system calculates a discharge strategy to fill the gap at the peak. To ensure executability, the strategy is verified for cross-sectional power flow and ramp rate after generation, and "next day's peak power" is reserved in cross-day scenarios. Finally, the suggested curves are implemented at the "nationwide—province—region—station" level: charging prioritizes low SOC stations, discharging prioritizes high SOC stations; independent energy storage is prioritized over paired energy storage; when regional targets are insufficient or exceeded, capacity is truncated and rolled back. The above process is repeated with a rolling window (e.g., 15 minutes) as forecasts are updated, ensuring consistency between the strategy and real-time status.

[0034] This invention revolves around four main lines: "standby capacity analysis—multi-timescale strategy generation—full-level target allocation—safety verification and closed-loop execution." The system first establishes a data channel with the SCADA / WAMS / planning and forecasting database at the dispatch master station, acquiring inputs such as the station cluster's SOC, capacity upper and lower limits, power upper limits, efficiency, and lockout status, as well as load, wind and solar power, power transmission, and the adjustable capacity of conventional units, all based on a unified time benchmark. Before data entry, the platform performs consistency verification on missing and outlier values, and then forms a rechargeable / deployable "capacity surface" at the provincial and network-wide levels, providing a baseline for subsequent strategy calculations and human-machine interaction. This preparation phase ensures that subsequent strategies are synchronized with real-time status and forecasts, and respect the station cluster's boundary conditions.

[0035] In terms of strategy generation, the system distinguishes between two scenarios: "new energy peak shaving obstruction" and "peak balance gap." When the midday new energy forecast significantly exceeds the load and external transmission demand, and conventional units are already close to their minimum output, the system calculates the obstruction and provides a corresponding charging curve. When a supply-demand gap occurs during the morning and evening peak periods, the system fills the gap with the peak of the discharge curve. To ensure executability and economy, the strategy undergoes cross-sectional power flow constraints and power ramp-up rate verification after generation, while also considering the reserved power for the "next day's peak" in the cross-day rolling scenario. In actual operation, the strategy is continuously updated with rolling windows of 15 minutes to adapt to real-time changes in forecasts and status. This unified "intra-day—pre-day—multi-day" process is the key to this invention's tight coupling of energy storage strategy and system operation.

[0036] Once the regional-level recommendation curve is formed, the system enters the hierarchical allocation and implementation phase of the targets. Allocation follows a hierarchy of "nationwide—province—region—station," prioritizing stations with lower State of Charge (SOC) during charging and stations with higher SOC during discharging. When both independent and integrated energy storage exist, priority is given to ensuring the scheduling space for independent energy storage. If regional targets are insufficient or exceeded, the system truncates and rolls back according to capacity and constraints, ensuring the allocation results are feasible within the station-level power upper limit, energy upper and lower limits, and ramp-up constraints. Simultaneously, to avoid conflicts between the strategy and the overall network targets, the system adopts a rule-based "global optimality" principle: no network-wide discharging is scheduled when renewable energy is obstructed; no charging is scheduled when there is a network-wide shortfall and no section overload occurs; and charging is not scheduled on non-section sides when conventional units have not reduced to minimum technical output. When a section overload occurs and does not affect network-wide support, on-site charging within the section is permitted to alleviate local congestion. These rules ensure both the physical feasibility of the strategy and the priority between renewable energy consumption and peak supply assurance.

[0037] Cross-day and cross-regional energy sharing are achieved through rolling forecasts and reserve capacity management. For example, if solar power is abundant during midday for two consecutive days while wind power is staggered during the evening peak, the system will release as much available energy as possible during the evening peak on the first day to reach the peak, and then resume charging on the second day, thus smoothing out source-load uncertainties over a longer time scale. The reverse scenario is executed according to the rhythm of "reserve capacity - discharge - mutual support". All cross-day strategies are re-evaluated within the rolling window based on the latest forecasts to avoid energy imbalances caused by forecast deviations.

[0038] Before the strategy is issued, safety verification and feasibility checks must be completed. The verification process, combining section limits and overloaded branches, simulates power flow changes caused by energy storage charging and discharging behavior. If voltage overload, branch overload, or section overload occurs, the system will adjust the plan in a "reduction-shift-redistribution" sequence until global and local constraints are met. Through this closed loop, the strategy can be implemented effectively while maintaining a safety margin during operation.

[0039] At the presentation and operation and maintenance level, the human-machine interface uses a four-layer perspective—panoramic / provincial / regional / station—to display the operational status and strategy comparisons. It also statistically analyzes indicators such as peak shaving power, curtailment rate, supply-demand gap reduction, charge / discharge ramp-up, and cycle costs. The system also supports offline strategy scoring and historical playback, allowing for the evaluation of strategy quality and the solidification of parameters and experience during operation and maintenance. These functions ensure the transparency and auditability of strategy operation.

[0040] Specifically, it includes the following steps: (1) Analysis of energy storage backup capacity (within the operating period T) Collect operational status data for each energy storage station (charge / discharge status, SOC, capacity, upper and lower limits of SOC, adjustment dead zone, etc.).

[0041] Calculate the station-level chargeable / dischargeable capacity based on the operational status data of each energy storage station: Rechargeable capacity: ; This formula describes the maximum energy reserve that a single station can continue to charge at the current time t. Real-time energy (MWh) The upper limit of energy (MWh). A safety dead zone (MWh) is reserved at both ends to avoid the protection / efficiency degradation band near the SOC upper limit. Results (MWh) is proportional to "the available space at the distance limit".

[0042] Dischargeable capacity: ; This formula describes the amount of energy that a single station can safely release at the current time t. A dead zone is also set for the lower energy limit (MWh). To avoid triggering lifetime and protection thresholds due to deep discharge. Result The larger the (MWh) value, the more abundant the "energy ammunition" available for peak performance.

[0043] The calculated station-level chargeable / dischargeable capacity is aggregated to the provincial and network-wide levels to obtain the chargeable / dischargeable capacity surface, which is then stored in the database to support human-machine interaction and subsequent strategy calculations.

[0044] (2) Multi-timescale strategy generation (day-to-day / 3-day / 5-day / intra-day rolling) Figure 2 The business process diagram generated for the multi-timescale strategy organizes the data flow according to the time sequence of "same day / multi-day - intraday rolling". First, a regional-level suggested curve is generated, then cross-section and ramp-up verification is performed, and the peak power of the next day is reserved in the cross-day scenario, and finally it enters the station group allocation. This process reflects the characteristics of the strategy being refined from top to bottom and continuously corrected in the rolling window.

[0045] Specifically, it obtains load forecasts at various scales, new energy power generation plans, AC / DC power transmission plans, and the maximum / minimum adjustable capacity of conventional units, and defines time sets. Station Collection ; regional set .

[0046] Retrieving station-level variables: These are the charging / discharging power and energy of station-level S, respectively. A single station cannot be charged and discharged simultaneously at the same time to match the mutual exclusion logic of the PCS / protection. The subscript 's' indicates the station-level identifier. The connection between this rule and the regional-level target is accomplished through the sorting and allocation of "region first, then station group" (charging is ordered in ascending order of SOC%, discharging is ordered in descending order of SOC%).

[0047] Constraints are set on the charging / discharging power and energy of station-level S:

[0048] ; in, , These are the upper limits of the charging / discharging power of station-level s, respectively. , These are the upper and lower limits of the energy at station level s, respectively. Let be the power change of station s between adjacent time points. This represents the upper limit of the slope within the station.

[0049] The above are the power upper limit, energy upper and lower limits, and ramp limit at the station level. When the recommended curves at the regional / network level are decomposed to the station level, these inequalities need to be checked station by station. If the limits are exceeded, they should be truncated and redistributed according to capacity.

[0050] Furthermore, an objective function based on the trade-offs of power curtailment / shortage / cycle cost is designed:

[0051] This goal is in the entire time domain. The internal parameters are: minimizing curtailment (Curt), positive gap (Gap+), and cycle cost. The weighted sum. The weights w1, w2, w3 can be given by the running policy preferences. This demonstrates the impact of cycle depth and number of cycles on lifetime depreciation.

[0052] in: ; ; ; .

[0053] Where Curt(t) represents the remaining amount of abandoned electricity (MW) at time t, defined as the positive part of the theoretically blocked peak-shaving power of new energy after deducting the current charging absorption. The larger Curt(t) is, the more new energy is blocked that has not been absorbed by the strategy during this period; the total time-domain aggregate is denoted as (MWh); This represents the remaining supply-demand balance gap (MW) at time t, defined as the positive portion of the system peak gap after deducting energy storage discharge compensation. The larger the value, the more likely there is a supply-demand gap that is not covered by the strategy during that period. The total time-domain aggregate volume is denoted as... (MWh); The cyclic equivalent cost at time t represents the lifetime depreciation and operational losses of electrochemical energy storage due to charge-discharge cycles. For ease of engineering implementation, this invention employs a linear throughput model: ,in , Δt represents the unit charging / discharging throughput cost coefficient (currency / MWh) for station-level s, and Δt is the time step (h). This model is compatible with linear objective functions in intraday rolling scenarios, making it easy to solve quickly. , , These are the predicted values ​​for new energy power, predicted system load power, and transmitted power (MW). , These represent the total output (MW) of conventional generating units at their operational limits during off-peak / peak periods. , These represent the charging / discharging power at all station levels.

[0054] Specifically, PRE stands for wind / solar power forecast (MW). The forecasting method is as follows: hourly forecasts are generated using the baseline method of "numerical weather forecast (NWP) + power curve of the power station" for the day-ahead / multi-day forecasts. During the day, the deviation between "measured power output and meteorological factors" is corrected by Kalman / exponential weighting in a rolling window of about 2-3 hours to ensure rolling updates at the 15-minute level. The forecast data is time-aligned, missing data is imputed, and out-of-bounds data is removed before being stored in the database.

[0055] PLoad is the system load power prediction (MW). The prediction method is as follows: a regression / tree model is established based on historical load, meteorological and calendar factors to form an hourly baseline, and the prediction is rolled over in 15-minute windows according to the latest measured deviation.

[0056] PExport is the outbound delivery plan curve (MW), which is derived from the scheduling plan and interconnection channel agreement.

[0057] The above-mentioned measurement of curtailment and deficit as "the remaining amount after the strategy is reduced" is more in line with the engineering requirement of "the effectiveness of the strategy can be verified".

[0058] Based on the above objective function and constraints, the target curves for two control scenarios are obtained, including the obstruction of new energy peak shaving (large generation of new energy at noon, low load) and the grid balance gap (peak gap in the morning and evening).

[0059] Specifically, the solution method and steps for the target curve are as follows: The above objective function and constraints can be solved within a unified optimization framework, specifically including: S1 Data Input: Loading theoretical peak-shaving power of new energy sources that is hindered according to a unified time base. The parameters include peak-segment power gap (Gap(t)), load / export / conventional unit capacity, station-level power and energy boundaries, ramping constraints, and efficiency parameters; and weights w1, w2, w3 and station-level s are set as unit charging / discharging throughput cost coefficients. , .

[0060] S2 decision variable: charging / discharging power of station s , and energy And introduce non-negative auxiliary variables including the cyclic equivalent cost at time t. The remaining amount of the gap needs to be balanced. This represents the remaining positive portion of the abandoned power and the power shortfall.

[0061] S3 constraint set: Including energy balance ,in, , These represent the charge / discharge efficiencies, ranging from (0,1]. , These represent the charging / discharging efficiencies of station-level s, respectively; this discrete energy balance equation maps charging and discharging power to energy trajectories. If only charging occurs during a certain period ( =0), energy according to Growth; if only discharge ( =0), energy according to Decrease. In the formula, energy is in MWh, power is in MW, and Δt is in h.

[0062] It also includes power / energy boundaries and ramping constraints; mutual exclusion logic (implemented using bigM + 0 / 1 variables). , Logical parameters ∈{0,1}); Also includes the , Linear definition: ; .

[0063] S4 objective function: .

[0064] S5 Solver and Scenarios: —Daily / Multi-day Scenario: Use Mixed Integer Linear Programming (MILP) to solve the problem (such as commercial or open-source solvers like CPLEX, Gurobi, HiGHS, etc.) to obtain globally feasible baseline charge / discharge targets and next day peak retention predictions; —Intraday rolling scenario: While maintaining the consistency of constraints, two methods can be used: ① Inherit MILP and resolve with a rolling window; ② To take into account timeliness, replace the mutual exclusion logic with soft constraints / large penalty coefficients, degenerate into linear programming (LP) or convex programming for fast recalculation (or use the heuristic fast reallocation of "charge first and then release / release first and then charge" as a failure alternative) to ensure 15-minute cascaded updates.

[0065] S6 Results Output and Verification: Obtain the network-wide target curves for two control scenarios (new energy peak shaving obstruction → charging target; grid balance gap → discharge target); then perform cross-sectional power flow and ramp verification, cross-day "next day peak" reservation, and station cluster decomposition implementation.

[0066] Figures 3 to 6 Key curves are provided for various aspects, including the theoretical peak-shaving obstruction of new energy sources, energy storage charging recommendations, reserve curves / balance gaps, and energy storage discharging recommendations. The obstruction curve defines the periods and magnitudes of absorption difficulties; the charging / discharging recommendation curve reflects the constraints of equipment and energy boundaries on the strategy; and the reserve curve depicts the time and scale of peak gaps. Together, these four curves constitute the measurement and execution basis of the strategy. The illustrations help maintenance personnel understand why the strategy chooses "charging without discharging" or "discharging without charging" during certain periods.

[0067] Specifically, the theoretical peak-shaving resistance curve for new energy sources is as follows:

[0068] This quantity represents the "hindered power" (MW) when the output of new energy sources exceeds the system's absorption capacity during off-peak hours. Among them, P... RE P represents the predicted power output of new energy sources such as wind and solar power. Load P is the predicted value of system load power. Export For external power transmission, This represents the minimum total output that a conventional unit can reduce to. If the value in parentheses is negative, it is counted as 0, indicating that there is no obstruction. The amount of obstruction directly affects the magnitude and timing distribution of the charging target.

[0069] In scenarios where peak shaving by renewable energy sources is hindered, a charging target is generated, and the recommended charging power is:

[0070] The recommended charging power (MW) is subject to three constraints: firstly, the upper limit of the "resistance power" C te Second, the upper limit of equipment charging power. Third, the upper bound of the power that the upper limit energy space can absorb within the current step size Δt. Efficiency η ch ∈(0,1] reflects the energy loss during the charging process.

[0071] Furthermore, the target curve for the backup curve / balance gap is:

[0072] This quantity represents the "power gap" (MW) caused by insufficient system supply capacity during peak hours. This represents the maximum total output that a conventional unit can achieve during peak hours. The shortfall directly drives the discharge target.

[0073] In a power grid imbalance scenario, a discharge target is generated, with a suggested discharge power of:

[0074] The recommended discharge power (MW) is also subject to three constraints: first, the upper limit of the "gap power"; second, the upper limit of the equipment discharge power. Third, the upper bound of the power that can be released within the current step size in the lower limit energy space. ηdis represents the discharge efficiency.

[0075] Furthermore, the cross-sectional power flow and ramp rate are verified for the peak-shaving obstruction curve and the target curve of the reserve curve / balance gap of the new energy theory.

[0076] Power / Energy / SOC Boundaries, Ramp Constraints: This constraint imposes a rate of change limit R (MW / step) on the proposed power trajectory between adjacent time periods, preventing the strategy from creating unenforceable "spikes" for AGC / unit ramp-up; In addition, the strategy at the station level must also meet boundary conditions such as charge / discharge blocking signals, cross-sectional power flow / heavy load limits, and AGC / AGC zone access. .

[0077] Furthermore, reserve "peak power for the next day" in cross-day scenarios: Recently—For several days: Based on forecasting and adjustability, a baseline curve is formed and the "peak" power for the next day is reserved; Specifically, "forecast" here refers to the steps involved in generating external inputs, including day-ahead / multi-day baseline forecasts and intraday rolling corrections for system load power forecasts (PLoad) and renewable energy power forecasts (PRE), as well as power transmission (PExport) and the lower limit of conventional unit capacity. The loading of adjustable capacity includes charge / discharge capacity surfaces, total output of conventional units, and cross-sectional power flow limits. This step is not equivalent to solving the objective function. After loading the above-mentioned prediction / plan and capacity boundaries, the system solves the optimization model in the same time domain as the prediction and adjustable capacity (MILP can be used for day-ahead / multi-day scenarios, and MILP rolling or degraded LP can be used for intraday scenarios for rapid recalculation), thereby forming the baseline charge / discharge curve and the "next day peak" reserve, and then proceeds to cross-sectional / ramp verification and station group decomposition. The baseline charge / discharge curve refers to the hourly charge / discharge power reference curve used to guide intraday rolling optimization in the day-ahead / multi-day time domain, which includes the time period and power allocation information for reserving the next day's peak power in cross-day scenarios.

[0078] Intraday: Update forecasts and status using a rolling window (e.g., every 15 minutes), resolv or quickly reallocate to ensure real-time executability and cost-effectiveness.

[0079] Cross-day / cross-regional mutual assistance (example): Strong solar power at noon and wind power peak in the evening on two days: discharge as much as possible during the evening peak on the first day and resume charging on the second day; or the opposite situation is implemented in a rolling manner according to the "reserved capacity - discharge - mutual assistance" strategy.

[0080] (3) Target allocation and coordination at all levels Figure 7 The diagram presents the implementation structure of target allocation, illustrating how regional targets are aligned with station-level capabilities and SOC order, while also coordinating with mutual exclusion, ramp-up, and cross-sectional constraints. The diagram emphasizes the "region first, then station cluster" implementation path, reflecting the recursive logic from aggregation to decomposition in engineering implementation. Key interfaces and data tables involved in the diagram are provided below.

[0081] Four-level collaboration: network-wide, province-level, region-level, and station-level. Network-wide target curve generation: Based on input data and global rules, generate the overall charge and discharge target curve for the entire network or a large region. Input data includes the network-wide target curve for theoretical peak shaving obstruction of new energy sources and grid balance gap, charge / discharge capacity and equipment / energy boundaries. The input data is constrained by conditions such as cross-section / tie line limits, power ramp-up, and charge / discharge mutual exclusion.

[0082] Following the principle of "global optimization" scheduling: when new energy sources are blocked, the entire network will not discharge; when the entire network is short-circuited (non-section heavy load), charging will not be carried out; when thermal power plants have not reached their minimum technical output, charging in non-section areas will not be arranged; when sections are heavily loaded, on-site charging in sections is allowed without affecting the support of the entire network.

[0083] Provincial Breakdown: The overall network target is broken down into provincial targets based on the proportion of energy storage capacity in each province and provincial section constraints.

[0084] Zonal decomposition: The provincial targets are refined into zonal targets based on zonal constraints (such as the quota of cross sections within a zonal area) and local capabilities (such as the number and type of energy storage stations within a zonal area).

[0085] Station-level allocation: Ultimately, the partition targets are precisely allocated to individual stations. Charging allocation: Stations are sorted by SOC% from smallest to largest, with priority given to fully charging stations with low SOC, while adhering to regional targets and station limits; Discharge allocation: Sorted by station SOC% from largest to smallest, with priority given to stations with high SOC for peak discharge; Independent energy storage takes precedence over integrated energy storage; when there is insufficient or overloaded capacity, it should be cut off according to capacity and constraints.

[0086] (4) Monitoring, evaluation and closed loop The human-machine interface (HMI) provides panoramic / provincial / regional / station operation monitoring and strategy comparison, and statistics on key indicators (peak shaving power, curtailment rate, supply and demand balance gap, charging and discharging ramp-up, cycle cost, etc.). Strategy simulation and safety verification: Verify cross-section support and over-limit passages to generate executable instructions or AGC targets; if requirements are not met, roll back according to "reduction-shift-redistribution" to maintain feasible operating conditions; Database storage: regional rechargeable / dischargeable meter, station-level policy table, etc., supporting traceability and auditing.

[0087] Example 2 This invention also proposes a multi-timescale energy storage charging and discharging monitoring and auxiliary decision-making system, comprising: 1) Data and Model Layer (Input Integration and Capability Surface Generation); a. Data Acquisition and Alignment: Establish data channels with SCADA / WAMS, planning and forecasting databases, and EMS relational databases; load station group SOC, capacity upper and lower limits, power upper limit, efficiency and blocking status, as well as load (PLoad), wind and solar (PRE), transmission plan (PExport), and conventional unit adjustable capacity (upper / lower limits and ramp-up) according to a unified time base; perform timestamp alignment, missing data imputation, and out-of-bounds removal on the original sequence. Output "cleaned time series data".

[0088] b. Model and parameter library: Maintain the basic / grid-connected / dynamic information model (ES_INFO) of energy storage equipment and parameters such as cross-section / tie line limits, as the boundary basis for subsequent strategy calculation and verification.

[0089] c. Capability Surface Calculation and Database Storage: Calculate the rechargeable / deployable energy reserve for each station and aggregate it into "rechargeable / deployable capability surfaces" by province and the entire network. Store this data in the AREA_ELEC_RET table (including timestamp and version number) for sharing between the business and presentation layers. Simultaneously, retain detailed information on available capabilities at the station level for subsequent layered allocation. The commonly used interface getCharADisCharCap() is responsible for aggregation and database writing.

[0090] 2) Business Layer (Strategy Generation - Verification - Decomposition) a. Multi-timescale strategy generation: Within a unified time series of "day-to-day / multi-day - intraday rolling", capacity surface and forecast / planned data are read. First, "Renewable Energy Obstacles → Charging Recommendation Curve" and "Peak Segment Gap → Discharge Recommendation Curve" are generated at the regional / network-wide scale to form baseline targets. The commonly used interface handleESCurve() outputs the recommendation curves for each region and scale. The results are stored in the database as table ES_POLICY_RET.

[0091] b. Safety and Feasibility Verification: Verify the cross-sectional power flow and power ramp rate of the proposed curve; if an over-limit is triggered, roll back according to the "reduction-shift-redistribution" method until the cross-sectional limit and unit ramp rate boundary are met. After passing, an "executable network-wide target curve" is formed.

[0092] c. Management of "next day peak" reserve capacity across days: In cross-day scenarios, power reserves and allocation are made based on the demand of the next day's peak period and the charging and discharging schedule of the day to avoid energy mismatch; this reserve capacity is updated on a rolling basis throughout the day.

[0093] d. Full-level target decomposition: According to the allocation path of "whole network - province - region - station", following the SOC sorting and station power / energy / mutual exclusion / ramp constraints, the regional target is refined into station-level target power and written into ESS_STRATEGY_RET; independent energy storage takes priority over attached energy storage, and when there is insufficient target / overload, it is truncated according to capacity and constraints.

[0094] III) Presentation Layer (Monitoring—Evaluation—Tracing) a. Operation monitoring and strategy comparison: The HMI provides a panoramic / province / region / station multi-dimensional perspective, displaying real-time status and strategy execution, and statistically analyzing indicators such as peak shaving power, curtailment rate, supply and demand gap reduction, charging and discharging ramp-up, and cycle cost.

[0095] b. Historical playback and scoring: Reads AREA_ELEC_RET, ES_POLICY_RET, ESS_STRATEGY_RET and verification logs, supports historical tracing and offline scoring; can perform offline evaluation of strategy feasibility and performance based on WOE / Logit and other methods and store the results in the database for parameter solidification and reporting.

[0096] (iv) Inter-layer interaction and runtime sequence (general technical roadmap) a. T0 Data Import: Data and model layers are cleaned and aligned → written to the input library and capability surface (AREA_ELEC_RET). b. T1 strategy generation: The business layer calls handleESCurve() to form a regional baseline suggestion → writes it to ES_POLICY_RET; c. T2 verification rollback: Perform section / slope verification → if necessary, “reduce-shift-redistribute” → generate executable region curve; d.T3 hierarchical allocation: Decompose the regional curve into station-level targets → write to ESS_STRATEGY_RET; e.T4 Display and Closed Loop: The display layer reads the above database tables and logs → displays / rates / reports; within the day, the T1–T4 cycle is triggered by scrolling the window to keep it synchronized with the forecast / status.

[0097] (v) Consistency between runtime environment and interface (corresponding to method) The system runs on a Linux+D5000 environment on the main station side, with the backend implementing core computing in C / C++ and the frontend in Java; external interfaces include SCADA / WAMS, planning and forecasting libraries, section / connection line quota and EMS relationship libraries; and a layered writing library for policy, execution and verification logs, supporting online closed loop and traceability.

[0098] This invention's system runs on a Linux+D5000 environment at the main station, with the backend using C / C++ for core computing and Java for the human-machine interface. External interfaces include: real-time SCADA / WAMS channels, load / wind / solar / transmission data from the planning and forecasting database, quota parameters for cross sections and tie lines, and read / write access to the EMS relational database. To meet traceability and auditing requirements, the system writes information such as regional chargeable / dischargeable capacity, suggested curves for each time scale, and station-level target power and status into the relational database. The "Regional Power Result Table (area_elec_ret)" records the capacity and timestamps for each province and the entire network, while the "Energy Storage Station Strategy Table (ess_strategy_ret / es_policy_ret)" records the station-level target power, time window, and execution status. Commonly used software interfaces include getCharADisCharCap() for capacity aggregation and database writing, and handleESCurve() for generating control targets for each region and time dimension.

[0099] Key implementation points of this invention (refer to the interface / function) include: Backup capabilities: getCharADisCharCap() summarizes data by province that can be added / added and written to the library; Strategy generation: handleESCurve() generates control targets for each region and scale; Data structures: ES_INFO, ES_POLICY_RET, etc.; Tables: AREA_ELEC_RET (regional electricity results), ESS_STRATEGY_RET / ES_POLICY_RET (policy results).

[0100] The multi-timescale energy storage charging and discharging monitoring and auxiliary decision-making method and system proposed in this invention realizes integrated strategy generation and highly reliable implementation across levels and time domains. Through unified suggestion curves, globally optimal scheduling rules, and station cluster allocation mechanisms, it can balance renewable energy consumption, supply and demand balance, and section safety in complex scenarios where renewable energy is hindered and peak-hour shortages coexist, demonstrating significant engineering feasibility and promotional value.

[0101] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit it. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that modifications or equivalent substitutions can still be made to the specific implementation of the present invention. Any modifications or equivalent substitutions that do not depart from the spirit and scope of the present invention should be covered within the protection scope of the claims of the present invention.

[0102] To make the objectives, technical solutions, and advantages of this invention clearer, the technical solutions of this invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of this invention. The embodiments described in this application are merely some embodiments of this invention, and not all embodiments. Based on the spirit of this invention, all other embodiments obtained by those skilled in the art without creative effort are within the protection scope of this invention.

[0103] This disclosure can be a system, method, and / or computer program product. A computer program product may include a computer-readable storage medium having computer-readable program instructions loaded thereon for causing a processor to implement various aspects of this disclosure.

[0104] Computer-readable storage media can be tangible devices capable of holding and storing instructions for use by an instruction execution device. Computer-readable storage media can be, for example—but not limited to—electrical storage devices, magnetic storage devices, optical storage devices, electromagnetic storage devices, semiconductor storage devices, or any suitable combination of the foregoing. More specific examples (a non-exhaustive list) of computer-readable storage media include: portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), static random access memory (SRAM), portable compact disc read-only memory (CD-ROM), digital multifunction disc (DVD), memory sticks, floppy disks, mechanical encoding devices, such as punch cards or recessed protrusions storing instructions thereon, and any suitable combination of the foregoing. The computer-readable storage media used herein are not to be construed as transient signals themselves, such as radio waves or other freely propagating electromagnetic waves, electromagnetic waves propagating through waveguides or other transmission media (e.g., light pulses through fiber optic cables), or electrical signals transmitted through wires.

[0105] The computer-readable program instructions described herein can be downloaded from computer-readable storage media to various computing / processing devices, or downloaded via a network, such as the Internet, local area network, wide area network, and / or wireless network, to an external computer or external storage device. The network may include copper transmission cables, fiber optic transmission, wireless transmission, routers, firewalls, switches, gateway computers, and / or edge servers. A network adapter card or network interface in each computing / processing device receives the computer-readable program instructions from the network and forwards them to the computer-readable storage media in the respective computing / processing device.

[0106] Computer program instructions used to perform the operations of this disclosure may be assembly instructions, instruction set architecture (ISA) instructions, machine instructions, machine-dependent instructions, microcode, firmware instructions, status setting data, or source code or object code written in any combination of one or more programming languages, including object-oriented programming languages ​​such as Smalltalk, C++, etc., and conventional procedural programming languages ​​such as the "C" language or similar programming languages. The computer-readable program instructions may execute entirely on the user's computer, partially on the user's computer, as a standalone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving a remote computer, the remote computer may be connected to the user's computer via any type of network—including a local area network (LAN) or a wide area network (WAN)—or may be connected to an external computer (e.g., via the Internet using an Internet service provider). In some embodiments, electronic circuitry, such as programmable logic circuitry, field-programmable gate arrays (FPGAs), or programmable logic arrays (PLAs), is personalized by utilizing the status information of the computer-readable program instructions to implement various aspects of this disclosure.

[0107] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit it. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that modifications or equivalent substitutions can still be made to the specific implementation of the present invention. Any modifications or equivalent substitutions that do not depart from the spirit and scope of the present invention should be covered within the protection scope of the claims of the present invention.

Claims

1. A method for monitoring and assisting decision-making in energy storage charging and discharging across multiple time scales, characterized in that, include: Collect operational status data from each energy storage station and calculate the station-level chargeable / dischargeable capacity based on the operational status data of each energy storage station; The calculated station-level chargeable / dischargeable quantities are aggregated to the entire network to obtain the chargeable / dischargeable capacity surface. Acquire station-level variables, including charging / discharging power and energy; Based on the rechargeable / dischargeable capacity and the charging / discharging power and energy, constraints are set, including energy balance, power / energy boundary and ramping constraints, mutual exclusion logic constraints, remaining curtailment constraints, and remaining supply-demand balance gap constraints; an objective function based on the trade-off between curtailment / gap / cycle cost is designed. Based on the objective function and constraints, the following are obtained: the whole-grid target curve and charging suggestion curve considering the obstruction of new energy peak shaving at multiple time scales, and the whole-grid target curve and discharging suggestion curve considering the grid balance gap. The cross-sectional power flow and ramp rate are checked on the charging and discharging recommendation curves. The power flow changes caused by the energy storage charging and discharging behavior are simulated. If voltage exceeds the limit, branch overload or cross-sectional overload occurs, the whole-network target curve and charging / discharging recommendation curve for the new energy peak shaving obstruction and grid balance gap are adjusted according to the actual situation until the constraints are met. After the verification is completed, the peak power of the next day is reserved in the cross-day scenario. Based on the grid-wide target curve of the obstruction of new energy peak shaving and the grid balance gap, and the charging / discharging strategy of reserving the peak power for the next day, the charging and discharging strategy is allocated and executed in the order of grid-province-region-station.

2. The method for monitoring and assisting decision-making of energy storage charging and discharging under multiple time scales according to claim 1, characterized in that: The mutual exclusion logic is implemented using a large M + 0 / 1 variable to constrain the charging / discharging power; The objective function based on the trade-off between curtailment / shortage / circular cost is in the full time domain. Minimize the curtailment surplus Curt, the supply-demand balance gap surplus Gap+, and the cycle equivalent cost. The weighted sum; The remaining amount of abandoned electricity Curt, the remaining amount of supply and demand balance gap Gap+, are related to the predicted value of renewable energy power, the minimum total output that conventional units can reduce to, the maximum total output that conventional units can increase, the predicted value of load power, the power transmitted to other units, and the charging / discharging power of all stations. The weights are given by the operating strategy preferences.

3. The method for monitoring and assisting decision-making of energy storage charging and discharging under multiple time scales according to claim 1, characterized in that: Based on the objective function and constraints, the network-wide target curves for two control scenarios—namely, the obstruction of renewable energy peak shaving and the grid balance gap—are obtained. The specific steps are as follows: For day-to-day / multi-day scenarios, mixed integer linear programming (MILP) is used to solve the problem and obtain a globally feasible baseline charge / discharge target. For intraday rolling scenarios, while maintaining the consistency of constraints, we can inherit MILP and resolve it with a rolling window, or replace the mutual exclusion logic with soft constraints / large penalty coefficients, degenerating into linear programming LP or convex programming for fast recalculation. The solution yields the network-wide target curves for the obstruction of new energy peak shaving and the grid balance gap.

4. The method for monitoring and assisting decision-making of energy storage charging and discharging under multiple time scales according to claim 1, characterized in that: The network-wide target curve for the obstruction of peak shaving in the new energy theory is represented as follows: The sum of the predicted value of renewable energy power and the minimum total output that can be reduced to conventional units is calculated. The sum is then subtracted from the predicted value of system load power and the power transmitted to the outside to obtain the difference result. The larger of 0 and the difference result is used to characterize the theoretical peak-shaving obstruction curve of renewable energy. The target curve for the entire power grid balance gap is represented as follows: The sum of the predicted system load power and the transmitted power is calculated. The sum is then subtracted from the predicted new energy power and the maximum total output that can be raised by conventional units to obtain the difference result. The larger of 0 and the difference result is used to characterize the grid-wide target curve of the grid balance gap.

5. The method for monitoring and assisting decision-making of energy storage charging and discharging under multiple time scales according to claim 4, characterized in that: The method for obtaining the predicted value of the new energy power is as follows: The hourly forecasts are generated using the baseline method of numerical weather prediction (NWP) and power curves of power plants. Within the day, the deviation between measured power output and meteorological factors is corrected by Kalman / exponential weighting using a near-rolling window to ensure rolling updates. The predicted new energy power values ​​are time-aligned, missing data are imputed, and out-of-bounds data are removed before being entered into the database. The method for obtaining the predicted load power value is as follows: A regression / tree model was recently established using historical load, meteorological and calendar factors to form an hourly baseline, which is then rolled over within a set time window based on the latest measured deviation.

6. The method for monitoring and assisting decision-making of energy storage charging and discharging under multiple time scales according to claim 1, characterized in that: Based on the network-wide target curve indicating that peak shaving is hindered by the aforementioned new energy theory, the following recommended charging power is generated: The recommended charging power is subject to three constraints, including the upper limit of the blocked power (Cte); and the upper limit of the device's charging power. ; The upper bound of the power that the upper limit energy space can absorb within the current step size Δt is calculated; the minimum value among the three constraints is used as the suggested charging power. Based on the grid-wide target curve for the aforementioned grid balance gap, a recommended discharge power is generated: It is recommended that the discharge power also be subject to triple constraints, including the upper limit of the gap power; and the upper limit of the device discharge power. ; The lower limit energy space is the upper bound of the power that can be released within the current step size; the minimum value among the three constraints is used as the suggested discharge power.

7. The method for monitoring and assisting decision-making of energy storage charging and discharging under multiple time scales according to claim 1, characterized in that: The specific method for reserving peak electricity for the next day in cross-day scenarios is as follows: Recently—For several days: The rechargeable / dischargeable capacity is used as the adjustable capacity. Based on the prediction and adjustable capacity, the baseline curve and the reserved peak power for the next day are obtained. The forecasts include day-ahead / multi-day baseline forecasts and intraday rolling corrections of system load power forecasts and renewable energy power forecasts, as well as the loading of external power and the lower limit of conventional unit capacity. The adjustable capabilities also include cross-sectional power flow limit, minimum total output that can be reduced to by conventional units, and maximum total output that can be raised by conventional units; Based on the predictive and adjustable capabilities, MILP is used for day-ahead / multi-day charging and discharging, and MILP rolling or degraded LP is used for intraday charging and discharging. The model is solved in the same time domain as the predictive and adjustable capabilities to obtain the baseline charging / discharging curve and the reserved peak power for the next day in cross-day scenarios. The baseline charge / discharge curve refers to the hourly charge / discharge power reference curve used to guide intraday rolling optimization within the daytime / multi-day time domain, including the time period and power allocation information for reserving peak power for the next day in cross-day scenarios; Intraday: Update forecasts and status using a rolling window, resolvable or quickly redistributed.

8. The method for monitoring and assisting decision-making of energy storage charging and discharging under multiple time scales according to claim 1, characterized in that: The charging and discharging strategy is allocated and executed in the order of network-wide, province-by-region, and station-by-station. Network-wide: Generate the overall target charge / discharge curve for the entire network: Following the principle of global optimal scheduling, the entire network will not discharge when renewable energy is blocked; charging will not be carried out when the entire network is short-circuited; charging will not be arranged outside the section until thermal power reaches its minimum technical output; and local charging within the section is allowed when the section is heavily loaded, provided that it does not affect the support of the entire network. By province: The overall network target is broken down into provincial targets based on the proportion of energy storage capacity in each province and the constraints of provincial sections; Zoning: The provincial objectives are refined into zonal objectives based on zonal constraints and local capabilities; Station-level: Precisely assigning zone targets to individual stations: Charging allocation: Stations are sorted by SOC% from smallest to largest, with priority given to fully charging stations with low SOC, while adhering to regional targets and station limits; Discharge allocation: Sorted by station SOC% from largest to smallest, with priority given to stations with high SOC for peak discharge; Independent energy storage is preferred over integrated energy storage. When there is insufficient or overloaded target, truncation is performed based on capacity and constraints.

9. A multi-timescale energy storage charging and discharging monitoring and auxiliary decision-making system utilizing the method described in any one of claims 1-8, comprising a data and model layer, a business layer, and a presentation layer, characterized in that: The data and model layer loads the station group operation data as the raw sequence; the raw sequence is then aligned with timestamps, imputed for missing data, and removed from out-of-bounds data to obtain cleaned time-series data. Calculate the chargeable / dischargeable capacity surface; At the business layer, station-level variables are obtained, including charging / discharging power and energy; constraints are set based on the charging / discharging capability and the charging / discharging power and energy. Design an objective function based on the trade-offs of curtailment / gap / cycle cost; based on the objective function and constraints, solve for the network-wide target curve considering the obstruction of new energy peak shaving and the grid balance gap at multiple time scales, as well as its charging and discharging suggested curves; verify the cross-sectional power flow and power ramp rate of the suggested curves; in the cross-day scenario, reserve and allocate power according to the peak demand of the next day and the charging and discharging schedule of the day; allocate paths in the order of network-province-region-station, following the SOC sorting and the power / energy / mutual exclusion / ramp constraints within the station, and refine the regional target into the station-level target power; The display layer provides a multi-dimensional perspective of panoramic / province / region / station, showing the real-time status and strategy execution, and statistically analyzing peak shaving power, curtailment rate, supply-demand gap reduction, charge-discharge ramp-up, and cycle cost indicators; it supports historical traceability and offline scoring; and it performs offline evaluation of strategy feasibility and performance and stores the results in the database for parameter solidification and reporting.

10. The energy storage charging and discharging monitoring and auxiliary decision-making system under multiple time scales according to claim 9, characterized in that: The system runs on a Linux+D5000 environment on the main station side, with the core computing implemented in C / C++ on the backend and Java on the frontend; the external interfaces include SCADA / WAMS, planning and forecasting libraries, cross-section / connection line quotas and EMS relationship libraries.

Images