Power grid gradeability dynamic evaluation method and system considering source storage cooperation

Abstract

The invention relates to the technical field of power system operation and control, in particular to a power grid gradeability dynamic evaluation method and system considering source-storage collaboration, and the method comprises the steps: obtaining prediction data and actual measurement data of a power grid load and new energy power generation power in real time, and carrying out the preprocessing of the obtained data, based on a quantile regression method, a power grid climbing demand in a future time period is dynamically evaluated, a new energy climbing capability dynamic evaluation model and an energy storage climbing capability dynamic evaluation model are respectively constructed, the new energy model evaluates a downward adjustment capability, the energy storage model evaluates a bidirectional real-time climbing capability under a charge state constraint, and a power grid climbing capability dynamic evaluation model is constructed through a source storage collaborative optimization model. The comprehensive climbing capacity of the power grid is calculated in an aggregated mode and compared with the climbing demand, and adequacy assessment and risk early warning are achieved; according to the method, the problems of inaccurate assessment of the gradeability and insufficient utilization of novel adjustment resources in the high-proportion new energy power grid are solved, and the absorption capability and the operation safety of the power grid to the fluctuating resources are improved.

Description

Technical Field

[0001] This invention relates to the field of power system operation and control technology, and in particular to a dynamic evaluation method and system for grid ramping capability considering source-storage synergy. Background Technology

[0002] The increasing penetration of new energy sources such as wind and solar power in the power system, coupled with the randomness, volatility, and intermittency of new energy output, leads to sharp and frequent fluctuations in the net load of the power grid (the difference between the total system load and the output of new energy sources). This poses an unprecedented challenge to the power system's ramp-up capability. Insufficient ramp-up capability may result in serious consequences such as frequency overruns, load shedding, or even large-scale power outages.

[0003] Currently, the power grid dispatch center's assessment of ramp-up capability mainly focuses on traditional thermal power and hydropower. The assessment method is relatively static and fails to fully consider the characteristics and potential of new regulation resources such as new energy sources and energy storage. For new energy sources, their ramp-up capability is mainly reflected in the direction of rapidly reducing output, but this capability is often overlooked. For energy storage, although its response speed is fast, its ramp-up capability is time-varying and conditional due to its state of energy (SOC), and traditional methods are difficult to conduct accurate dynamic assessments of it.

[0004] Current technologies for assessing grid ramping capacity primarily focus on slow-regulating resources such as traditional thermal and hydropower. Their models are relatively static. Furthermore, existing methods for forecasting ramping demand largely rely on confidence intervals based on historical statistics, failing to fully utilize the nonlinear relationship between predicted values ​​and errors. This leads to decreased accuracy in scenarios with a high proportion of renewable energy, and inaccurate predictions when faced with the strong volatility and non-normal distribution of renewable energy output. Therefore, there is an urgent need for a method and system capable of dynamically and accurately assessing grid ramping capacity, especially one that fully considers the synergistic effect of power generation and storage to unlock the overall flexible regulation potential of the system and ensure the safe and stable operation of the power grid. Summary of the Invention

[0005] This invention provides a dynamic assessment method and system for grid ramping capacity considering source-storage synergy, which solves the problems of inaccurate ramping capacity assessment and insufficient utilization of new regulation resources in grids with a high proportion of new energy sources, and improves the grid's ability to absorb fluctuating resources and its operational safety.

[0006] To achieve the above objectives, the present invention employs the following technical solution: A dynamic assessment method for grid ramping capability considering source-storage synergy includes the following steps: S1. Real-time acquisition of predicted and measured data on grid load and new energy power generation, and preprocessing of the acquired data; S2. Dynamically assess grid ramping demand in future periods based on quantile regression method; S3. Construct dynamic evaluation models for the ramping capabilities of new energy and energy storage respectively. The new energy model evaluates its downward adjustment capability, while the energy storage model evaluates its bidirectional real-time ramping capability under state of charge constraints. S4. Through the source-storage collaborative optimization model, the comprehensive grid ramping capacity is aggregated and calculated, and compared with the ramping demand to achieve adequacy assessment and risk warning.

[0007] Furthermore, the predicted and measured data of grid load and new energy power generation include: real-time values ​​and ultra-short-term prediction sequences of total grid load, real-time output and minute-level or 15-minute-level power prediction values ​​of wind farms and photovoltaic power stations, meteorological telemetry data of wind speed, irradiance, and cloud cover, and real-time charging and discharging power, core state parameters (SOC, rated power, and energy capacity) of energy storage systems.

[0008] Furthermore, the real-time acquisition of predicted and measured data on grid load and new energy power generation comes from grid dispatch automation systems, wind farm power prediction systems, photovoltaic power prediction systems, meteorological information systems, and energy storage management systems.

[0009] Furthermore, step S2 specifically includes the following steps: S2.1 Model Training: Using historical data, quantile quadratic regression fitting is performed on the prediction errors and predicted values ​​of load, wind power, and photovoltaic power respectively to obtain the prediction error curve equations for each. S2.2 Differentiated processing: Based on the daytime characteristics of photovoltaic power generation, if the predicted value exceeds 50% of the maximum predicted value for the day, the data is classified as strong light; if the predicted value does not exceed 50% of the maximum predicted value for the day, the data is classified as weak light, and regression fitting is performed separately. S2.3 Real-time forecasting: Data on system load, wind power, and photovoltaic power generation from multiple historical operating days are collected, including predicted and actual values. The positive and negative errors in the system load and power generation forecasts for each time period are calculated using the following formula: ; ; ; in, This refers to the system load forecast error for time period t calculated based on historical data. , These represent the actual and predicted values ​​of the system load for time period t, respectively. The wind power prediction error for time period t is calculated based on historical data. , These represent the actual and predicted wind power output for time period t, respectively. The photovoltaic power generation prediction error for time period t is calculated based on historical data. , These represent the actual and predicted photovoltaic power generation for time period t, respectively.

[0010] Furthermore, the new energy model assesses its downward adjustment capability; the new energy ramp-up capability assessment model is as follows: ; in, It is the absolute physical upper limit that can be adjusted downwards; Lowering the boundary for safety; For downward adjustment capability; This is the predicted power output of the power station; A configurable confidence coefficient, set by the scheduling agency based on its risk tolerance for forecast uncertainty; This represents the standard deviation of the recent power prediction error for the power station.

[0011] Furthermore, the energy storage model evaluates its bidirectional real-time ramping capability under state of charge constraints. The energy storage ramping capability evaluation model is based on real-time available power calculation. ; ; in, , These are the available discharge power and available charging power of the energy storage device, respectively. S represents the rated energy storage capacity of the energy storage device; S represents the real-time energy state. , These are the upper and lower limits of the energy state that is allowed to operate; The rated energy capacity of the energy storage; The time required for the expected energy storage to continuously provide ramp-up capability.

[0012] Furthermore, the formula for calculating the overall grid ramping capability is as follows: ; ; in, , These refer to the upward and downward climbing capabilities of energy storage devices.

[0013] A dynamic evaluation system for grid ramping capability considering source-storage coordination includes: Data acquisition module: used to acquire real-time predicted and measured data of grid load and new energy power generation, and to preprocess the acquired data; Ramp-up Demand Assessment Module: Data acquired by the data acquisition module is input into the ramp-up demand assessment module, which dynamically assesses the grid ramp-up demand for future periods based on quantile regression. Climbing Capability Assessment Module: Used for dynamic assessment of climbing capability for various resource types; Source-storage collaborative aggregation module: used to aggregate and calculate the overall grid ramping capacity, compare the calculated overall grid ramping capacity with the ramping demand assessed in the ramping demand assessment module, and realize adequacy assessment. Visualized early warning module: Visualizes and provides risk warnings based on the results of the comparison by the source-storage collaborative aggregation module.

[0014] Compared with the prior art, the beneficial effects of the present invention are: 1) Improved assessment accuracy and reliability: By introducing quantile regression to assess ramp-up demand, this invention can accurately capture the tail risks in new energy output and load forecasting, significantly reducing risk misjudgment or overly conservative reserves caused by inaccurate demand forecasting. Compared with the traditional confidence interval method based on historical statistics, the accuracy of the assessment results is expected to be improved by more than 15%. By establishing a "new energy downward adjustment capacity model" and a "storage SOC constraint dynamic model", this invention changes the previous situation of "ignoring" or "simply exaggerating" the ramp-up capacity of new resources. The assessment results are highly consistent with the actual physical characteristics and real-time operating status of the resources, avoiding overestimation or underestimation of capacity. This allows dispatchers to rely on the assessment results for decision-making, thus greatly enhancing the reliability of system operation. 2) Enhanced System Flexibility and Renewable Energy Absorption Capacity: This invention, for the first time, systematically incorporates the rapid output reduction capability of new energy sources into the ramp-up supply side, providing the power grid with a new, low-cost, and rapidly responsive ramp-up adjustment approach. When facing steep downhill ramp events, new energy sources can be prioritized for output reduction, reducing reliance on traditional thermal power units and thus lowering the overall system regulation cost. Through a source-storage synergistic optimization model, this invention achieves optimal coordination of heterogeneous resources such as new energy and energy storage on a spatiotemporal scale. This synergistic effect fully taps into the system's potential flexibility, enabling the power grid to safely accommodate a higher proportion of new energy sources under the same hardware conditions, effectively reducing wind and solar curtailment rates. 3) Improved Dispatch Decision-Making and Risk Control Capabilities: The minute-level rolling assessment and forward-looking risk warning provided by this invention offer dispatchers valuable decision-making lead time. The system can identify the risk of insufficient ramping capacity in advance and clearly define the capacity shortage, giving dispatchers sufficient time to activate contingency plans, adjust methods, or call up ramping auxiliary services, realizing a shift from "passive response" to "proactive prevention." Through an intuitive and visual human-machine interface, the complex system status is transformed into a clear "dashboard" and "early warning signals," greatly reducing the cognitive load on dispatchers and supporting them to make rapid and accurate decisions. This not only improves the power grid's security defense level but also enhances the overall efficiency of dispatch operations. Attached Figure Description

[0015] Figure 1 This is a scatter plot of the system net load prediction error rate in an embodiment of the present invention.

[0016] Figure 2 This is a scatter plot of the system load prediction error rate in an embodiment of the present invention.

[0017] Figure 3 This is a schematic diagram of the deterministic and uncertain uphill climb curves in a typical daily embodiment of the present invention.

[0018] Figure 4 This is a schematic diagram of the gradient curves under deterministic and uncertain conditions for a typical day in an embodiment of the present invention. Detailed Implementation

[0019] The specific embodiments of the present invention will be further described below with reference to the accompanying drawings: The present invention provides a dynamic assessment method for grid ramping capability considering source-storage synergy, comprising: S1, acquiring real-time predicted and measured data of grid load and new energy power generation, and preprocessing the acquired data; First, the system needs to perform large-scale, parallel real-time data acquisition from multiple heterogeneous data sources, including power grid dispatch automation systems, new energy power prediction systems, meteorological information systems, and energy storage management systems. The acquired data covers all elements reflecting the system's supply and demand status, including: Load data: Real-time values ​​and ultra-short-term forecast sequences of total grid load; New energy data: real-time output of wind farms and photovoltaic power plants, high-precision minute-level or 15-minute-level power prediction values, and corresponding meteorological telemetry data such as wind speed, sunshine, and cloud cover; Energy storage system data: real-time charge and discharge power, core state parameters (SOC), rated power and energy capacity, operating status, etc. Data preprocessing: These raw data differ in time series, quality, and format, so they must undergo a rigorous preprocessing process to clean, align, interpolate, and normalize the collected raw data, eliminate outliers, ensure data consistency across time and space, form a standardized evaluation data pool, and ensure that all data are under a uniform timestamp and sampling frequency to eliminate the influence of units.

[0020] S2. Dynamically assess grid ramping demand in future periods based on quantile regression method; (1) Net load calculation: Calculate the system net load forecast for the current and future rolling time windows (with 15-minute intervals).

[0021] (2) Quantification of uncertain ramp-up demand: The prediction error of the system net load is estimated by using the quantile regression method, and this error is the uncertain ramp-up demand; Traditional methods based on historical statistical confidence intervals often fail when faced with non-Gaussian and non-linear fluctuations brought about by new energy sources. This invention, after obtaining high-quality data, uses quantile regression to focus on prediction errors, thereby more scientifically assessing the uncertainty ramp-up requirements. This method first uses massive historical data to model the relationship between the prediction errors and their predicted values ​​for system load, wind power, and photovoltaic power. It does not simply estimate the average error, but rather uses quantile quadratic regression to fit an upper boundary curve equation describing the error most likely to reach under a given prediction value. Furthermore, considering the drastically different fluctuation characteristics of photovoltaic power generation during "strong light" and "weak light" periods within a day, this method performs differentiated processing: based on whether the photovoltaic prediction value exceeds 50% of the maximum prediction value for the day, the data is divided into two subsets, and quantile regression is performed separately for each subset, thus obtaining a more accurate, state-related error model. In real-time assessment, the system substitutes the predicted values ​​of load, wind power, and photovoltaic power within the future rolling window into their respective quantile regression equations, and adds up the obtained error components to obtain the ramp-up demand capacity that the system must prepare to cope with uncertainties in the future period. This process dynamically and adaptively captures extreme scenarios of system net load fluctuations and provides accurate target values ​​for ramp-up capacity assessment. The specific method includes the following steps: S2.1 Model Training: Using historical data, quantile quadratic regression fitting is performed on the prediction errors and predicted values ​​of load, wind power, and photovoltaic power respectively to obtain the prediction error curve equations for each. S2.2 Differentiated processing: Based on the daytime characteristics of photovoltaic power generation, if the predicted value exceeds 50% of the maximum predicted value for the day, the data is classified as strong light; if the predicted value does not exceed 50% of the maximum predicted value for the day, the data is classified as weak light, and regression fitting is performed separately. S2.3 Real-time forecasting: Data on system load, wind power, and photovoltaic power generation from multiple historical operating days are collected, including predicted and actual values. The positive and negative errors in the system load and power generation forecasts for each time period are calculated using the following formula: ; ; ; in, This refers to the system load forecast error for time period t calculated based on historical data. , These represent the actual and predicted values ​​of the system load for time period t, respectively. The wind power prediction error for time period t is calculated based on historical data. , These represent the actual and predicted wind power output for time period t, respectively. The photovoltaic power generation prediction error for time period t is calculated based on historical data. , These represent the actual and predicted photovoltaic power generation for time period t, respectively.

[0022] S3. Construct dynamic evaluation models for the ramping capabilities of new energy and energy storage respectively. The new energy model focuses on evaluating its downward adjustment capability, while the energy storage model focuses on its bidirectional real-time ramping capability under state of charge constraints. After clarifying the system requirements, the ramp-up capability, or "supply potential," of various resources in the system is accurately assessed. This invention abandons the model of simply treating resources as fixed output units. Instead, it establishes differentiated and refined dynamic assessment models based on the physical nature and control characteristics of different types of resources. For renewable energy power plants, they are redefined as flexible resources with "negative value" adjustment capabilities. Wind and solar curtailment is regarded as a downward adjustment capability, and a model is constructed to quantify its rapid downward adjustment capability. The core of this model is to take the minimum value between the current actual output and a "safe downward adjustment boundary" that takes into account the uncertainty of prediction. This definition liberates renewable energy from the identity of a pure "ramp-up demand manufacturer" for the first time, and incorporates its ability to rapidly reduce output into the ramp-up supply side, providing a new and more economical adjustment means for dispatch. For energy storage systems, this invention profoundly grasps their essence as "energy resources," constructing a dynamic correlation model between their ramp-up capability and real-time State of Charge (SOC). By monitoring their energy state in real time and calculating the truly usable charging and discharging power of the system under the current state, combined with their rapid power response characteristics, it derives their currently usable bidirectional ramp-up capability that can be sustained for a certain period of time. Specifically, it includes the following: (1) Assessment of the ramp-up capability of new energy sources: The focus is on assessing their ability to adjust rapidly downwards. The model is as follows: In the formula: It is the absolute physical upper limit for downward adjustment; the power output of the power station cannot be reduced below zero. To safely lower the threshold, it means that, considering the uncertainty of forecasting, the actual output in the future may be higher than the current forecast. Therefore, not all current output can be simply regarded as adjustable capacity; a portion must be reserved. To address this upward prediction error and avoid violating dispatch instructions or exacerbating imbalances due to an increase in actual power after reducing output; For downward adjustment capability; This is the predicted power output of the power station; A configurable confidence coefficient, set by the scheduling agency based on its risk tolerance for forecast uncertainty; This represents the standard deviation of the recent power prediction error for the power station. Principle: Based on the uncertainty of prediction, this model quantifies the upper limit of output that new energy sources can safely and quickly reduce under dispatch instructions, transforming them from "problem creators" to "solution providers". (2) Energy storage ramp-up capability assessment: The focus is on assessing its bidirectional and rapid ramp-up capability under SOC constraints; Real-time available power calculation: ; ; in, , These are the available discharge power and available charging power of the energy storage device, respectively. S represents the rated energy storage capacity of the energy storage device; S represents the real-time energy state. , These are the upper and lower limits of the energy state that is allowed to operate; The rated energy capacity of the energy storage; The time required for the expected energy storage to continuously provide ramp-up capability is a system-level parameter used to calculate energy constraints. Downward adjustment capability = This quantitative model redefines new energy as a flexible resource with the ability to adjust rapidly downwards, and dynamically incorporates its capabilities into the ramp-up supply side, greatly expanding the scheduling resources. The source-storage synergistic aggregation optimization method for multiple types of resources is not a simple summation of the capabilities of various resources. Instead, it constructs a mathematical optimization model with the goal of minimizing costs or shortages, and uses ramp-up demand and the dynamic capabilities of various resources as constraints to solve for the globally optimal ramp-up capability reservation and allocation scheme, achieving a synergistic benefit of "1+1>2".

[0023] S4. Through the source-storage collaborative optimization model, the comprehensive grid ramping capacity is aggregated and calculated, and compared with the ramping demand to achieve adequacy assessment and risk warning; Dynamic climbing ability calculation: ; ; in, , These are the uphill and downhill ramp capabilities of the energy storage device, respectively. The ramp capability of energy storage is usually equal to its available power. This model ensures that the evaluated ramp capability is not an instantaneous pulse, but a capability that can be provided continuously and stably within the time scale required for scheduling. When the SOC approaches the upper limit, its downhill (charging) capability decreases sharply; when the SOC approaches the lower limit, its uphill (discharging) capability also decreases. This dynamic decay characteristic is the key to the present invention being able to truly reflect the operating status of energy storage.

[0024] A dynamic evaluation system for grid ramping capacity considering source-storage synergy, employing a combination of data-driven and model-driven approaches, and a unified approach of divide-and-conquer and collaborative aggregation, achieves a refined and dynamic evaluation of grid ramping capacity, including: Data acquisition module: used to acquire real-time predicted and measured data of grid load and new energy power generation, and to preprocess the acquired data; the output standardized, high-quality data pool provides reliable and consistent input for all subsequent advanced analyses, and is the solid data foundation of the entire evaluation system; Ramp-up Demand Assessment Module: Data acquired by the data acquisition module is input into the ramp-up demand assessment module. Within the ramp-up demand assessment module, the grid ramp-up demand for future periods is dynamically assessed based on the quantile regression method. It dynamically and adaptively captures extreme scenarios of system net load fluctuations, providing accurate target values ​​for ramp-up capacity assessment. Climbing Capability Assessment Module: Used for dynamic assessment of climbing capability for various resource types; Source-storage collaborative aggregation module: used to aggregate and calculate the overall grid ramping capacity, compare the calculated overall grid ramping capacity with the ramping demand assessed in the ramping demand assessment module, and realize adequacy assessment; effectively prevents the problem of inflated ramping capacity due to ignoring energy constraints, and ensures the physical feasibility of the assessment results; Visualized early warning module: Visualizes and provides risk warnings based on the results of the comparison by the source-storage collaborative aggregation module.

[0025] This invention addresses the fundamental problem of insufficient characterization of novel regulatory resources in existing assessment methods: it expands the assessment scope to include novel regulatory resources such as wind power, photovoltaics, electrochemical energy storage, and pumped storage, aiming to accurately characterize the "negative" ramp-up capability of new energy resources, i.e., their controllable and rapid downward regulation capability, changing the traditional perception that they are only regarded as "sources of ramp-up demand" while ignoring their "ramp-up supply potential." At the same time, it aims to dynamically quantify the "constrained" ramp-up capability of energy storage resources, establishing a real-time, dynamic correlation model between their strong bidirectional ramp-up capability and their internal core state—State of Energy (SOC), avoiding inflated capability assessments due to neglecting SOC constraints. Overcoming the shortcomings of poor accuracy in existing ramp-up demand forecasting in new energy scenarios: By introducing quantile regression, the nonlinear relationship between the system net load forecasting error and the forecast value itself can be captured more accurately. In particular, the differential fluctuation characteristics of photovoltaic power generation under different light intensities can be distinguished and modeled, thereby achieving a more refined and reliable quantitative assessment of uncertain ramp-up demand. Breaking down the barriers of isolated evaluation of multiple types of resources and realizing the synergistic benefits of source-storage synergy: In order to avoid simply treating new energy, energy storage and other resources as independent entities for capacity superposition, this invention constructs a source-storage synergy optimization model. Through this model, the complementary advantages and spatiotemporal optimization configuration of the ramp-up capabilities of heterogeneous resources can be realized.

[0026] Simulation results show that, compared to traditional methods, the following results are available (see Table 1): Table 1 Comparison of Simulation Verification Modeling Results project Confidence number statistics Quantile regression method Coverage / % 95.27 96.55 Positive error exceeding amount / MW 120.56 103.9 Positive error prediction / MW 24.3 26.5 Negative error excess / MW 112.5 102.9 Negative error prediction / MW 25.3 28.1 Based on historical operating day data, the system net load forecast error rates corresponding to confidence levels of 95%, 90%, and 85% were calculated and determined using the confidence number statistical method, and plotted as line graphs. Simultaneously, the system net load forecast error rate for each time period was plotted as a scatter plot, as shown below. Figure 1 As shown.

[0027] Based on historical operating day data, the prediction error curve equations for system load, photovoltaic power generation, and wind power at confidence levels of 95% (upper quantile 97.5%, lower quantile 2.5%), 90% (upper quantile 95.0%, lower quantile 5.0%), and 85% (upper quantile 92.5%, lower quantile 7.5%) were calculated using the quantile regression method and plotted as curves. Simultaneously, the prediction error for each time period was plotted as a scatter plot, as shown below. Figure 2 As shown.

[0028] The following embodiments are implemented based on the technical solution of the present invention, providing detailed implementation methods and specific operation processes. However, the scope of protection of the present invention is not limited to the following embodiments. Unless otherwise specified, the methods used in the following embodiments are conventional methods.

[0029] Example This invention selects data from 40 historical operating days in a certain province for processing and analysis. Based on the obtained system net load prediction error pattern, it estimates the prediction error of four typical days (weekdays, Saturdays, Sundays, and holidays) and evaluates their ramp-up ability. Then, it conducts a case study analysis on the obtained results. Based on spot electricity market clearing data from four typical days, the calculation results for deterministic ramp-up and uncertain ramp-up are as follows: Figure 3 and Figure 4 As shown.

[0030] The system loads selected in this section, from highest to lowest, generally follow a pattern of weekdays, Saturdays, Sundays, and holidays. Generally, the higher the system load, the more likely the energy storage devices are to reach their maximum output capacity, resulting in insufficient uphill ramping capacity but ample downhill ramping capacity; conversely, the lower the system load, the more likely the energy storage devices are to reach their minimum output capacity, resulting in insufficient downhill ramping capacity but ample uphill ramping capacity. Different regions employ different methods for predicting system load and renewable energy generation, and load and generation also have regional characteristics, leading to varying data characteristics of errors.

[0031] Analysis shows that the uncertainty ramp determined by the quantile regression method is generally greater than the uncertainty ramp determined by the confidence number statistical method, which is consistent with the calculation results of the estimated indicators in Table 1.

Claims

1. A method for dynamic evaluation of power grid ramping capacity considering source-storage synergy, characterized in that, It comprises the following steps: S1, real-time acquisition of power grid load, new energy power generation prediction data and measured data, and pre-processing of the acquired data; S2, dynamically evaluate the power grid climbing demand of the future period based on the quantile regression method; S3, respectively construct new energy and energy storage climbing ability dynamic evaluation model, wherein the new energy model evaluates its downward adjustment ability, and the energy storage model evaluates its real-time climbing ability under the state of charge constraint; S4, through the source and storage collaborative optimization model, aggregate the calculation of the comprehensive climbing ability of the power grid, and compare it with the climbing demand to realize the adequacy evaluation and risk warning.

2. The method of claim 1, wherein, The prediction data and measured data of the power grid load and new energy power generation include: real-time value and ultra-short-term prediction sequence of total power load, real-time output of wind farms and photovoltaic power stations, minute-level or 15-minute-level power prediction value, wind speed, illumination, cloud cover meteorological telemetry data, real-time charging and discharging power of energy storage system, core state parameter energy state SOC, rated power and energy capacity.

3. The method of claim 2, wherein the method further comprises: The source of the real-time acquisition of the power grid load, new energy power generation prediction data and measured data is the power grid dispatching automation system, wind farm power prediction system, photovoltaic field prediction system, meteorological information system and energy storage energy management system.

4. The method of claim 1, wherein, The step S2 specifically comprises the following steps: S2.1, model training: using historical data, respectively fitting the prediction error and prediction value of load, wind power and photovoltaic power by quantile quadratic regression to obtain the prediction error curve equation of each; S2.2, differential processing: for the daytime characteristics of photovoltaic power generation, the prediction value exceeding 50% of the daily maximum prediction value is divided into strong light, and the prediction value not exceeding 50% of the daily maximum prediction value is divided into weak light, and regression fitting is performed respectively; S2.3, real-time estimation: taking the system load, wind power and photovoltaic power data of multiple historical operation days, including prediction value and actual value, calculating the positive error and negative error of system load and power generation power prediction of each period, and the calculation formula is: ； ； ； wherein, is the system load prediction error for period t calculated based on historical data; , are the magnitude of the actual and predicted system load for period t, respectively; is the wind power prediction error for period t calculated based on historical data; , are the actual and predicted wind power for period t, respectively; is the photovoltaic power prediction error for period t calculated based on historical data; , are the actual and predicted photovoltaic power for period t, respectively.

5. The method of claim 1, wherein, The new energy model evaluates its downward adjustment ability, and the new energy climbing ability evaluation model is: ； wherein, is the absolute physical upper limit of down regulation; is the safe down regulation boundary; is the down regulation capability; is the site power forecast value; is the configurable confidence factor set by the dispatch authority according to the risk tolerance for forecast uncertainty; is the standard deviation of the site recent power forecast error.

6. The method of claim 1, wherein, The energy storage model evaluates its real-time climbing ability under the state of charge constraint, and the energy storage climbing ability evaluation model is real-time available power calculation: ； ； wherein, , are the available discharge power and the available charge power of the energy storage device, respectively; is the energy storage rated power of the energy storage device; S is the real-time state of energy; , are the upper and lower energy state limits for allowed operation, respectively; is the rated energy capacity of the energy storage; is the time for which the energy storage is expected to continuously provide ramping capability.

7. The method of claim 6, wherein the method further comprises: The aggregate calculation of the comprehensive climbing ability of the power grid is as follows: ； ； wherein, , are the upward and downward ramping capabilities of the energy storage device, respectively.

8. An evaluation system using the method for dynamic evaluation of the climbing ability of a power grid considering the synergy of sources and storages according to any one of claims 1-7, characterized in that, It comprises: Data acquisition module: used for real-time acquisition of power grid load, new energy power generation prediction data and measured data, and pre-processing of the acquired data; Climbing demand evaluation module: the data acquired by the data acquisition module is input into the climbing demand evaluation module, and the climbing demand of the future period is dynamically evaluated in the climbing demand evaluation module based on the quantile regression method; Climbing ability evaluation module: used for dynamic evaluation of climbing ability of multiple types of resources; Source and storage collaborative aggregation module: used for aggregate calculation of the comprehensive climbing ability of the power grid, and comparison of the calculated comprehensive climbing ability of the power grid with the climbing demand evaluated in the climbing demand evaluation module to realize adequacy evaluation; Visual warning module: the results after comparison of the source and storage collaborative aggregation module are visually displayed and risk warned.

Images