Flexible load resource adjustment method based on multi-power index combination weighting

Abstract

The invention discloses a flexible load resource adjustment method based on multi-power index combination weighting, and belongs to the technical field of load adjustment. According to the method, comprehensive modeling is carried out on flexible load resources of a power system through multiple power index sets and power indexes of multiple dimensions and multiple types; classifying the power indexes, and distinguishing the adjustment relevance of different power indexes in a flexible load resource adjustment scene according to the flexible load resource adjustment scene relevance type; the combination weight is determined through the scene adaptation weight and the power influence weight, and the influence of the flexible load resource adjustment scene and the power index on the operation of the power system is considered; weight optimization is carried out through a machine learning algorithm, so that the accuracy, rationality and scene adaptability of weight distribution are remarkably improved; therefore, flexible load resource adjustment can be performed by adjusting the scene weight and the flexible load resource adjustment scene association type, and the comprehensiveness and accuracy of flexible load resource adjustment are improved.

Description

Technical Field

[0001] This invention belongs to the field of load regulation technology, and in particular relates to a flexible load resource regulation method based on the weighting of multiple power index combinations. Background Technology

[0002] As the penetration rate of renewable energy in the power system continues to rise, the strong fluctuations and randomness of the output of renewable energy sources such as wind power and photovoltaics have exacerbated the power balance pressure in the operation of the power system, leading to increasingly tight regulation resources. Flexible load resources, as a core regulation resource of the new power system, play an irreplaceable role in ensuring the safety and stability of the power grid, improving the operational flexibility of the power system, and promoting deep interaction between power sources and loads. With its wide distribution, diverse types, and rapid response, flexible load resources can dynamically adjust electricity consumption behavior in scenarios such as peak shaving, frequency regulation, and reserve through regulation operations such as industrial interruptible loads, intelligent charging and discharging of electric vehicles, and cold and heat storage systems. This effectively mitigates the impact of renewable energy output fluctuations on the power grid and enhances the power system's ability to absorb renewable energy.

[0003] Current traditional schemes for flexible load resource regulation mostly rely on single power indicators, such as maximum adjustable capacity or response time, or simply static modeling of power system operation. This one-sided and static modeling approach fails to comprehensively depict the integrated regulation capabilities of flexible loads under complex operating conditions. Consequently, flexible load resource regulation cannot fully reflect the actual regulation needs of the power system and cannot accurately match the actual operating conditions of the power system. Furthermore, the flexible load resource regulation scenarios in power systems vary greatly, such as peak shaving and valley filling, frequency regulation, spinning reserve, and black start, leading to significant differences in the regulation requirements for load resources. One-sided and static modeling methods cannot accurately regulate flexible load resources. Therefore, there is an urgent need for a flexible load resource regulation method based on a combination of multiple power indicators with weighting to address the shortcomings of existing technologies. Summary of the Invention

[0004] This invention aims to provide a flexible load resource regulation method based on the weighting of multiple power indicators to solve the above-mentioned technical problems. By constructing a set of multiple power indicators and combining weights, the accuracy and comprehensiveness of flexible load resource regulation are improved.

[0005] To address the aforementioned technical problems, embodiments of the present invention provide a flexible load resource adjustment method based on a combination of multiple power index weights, comprising: Acquire real-time operational data of the power system and flexible load resource adjustment scenarios; Based on the real-time operating data, a load regulation index set, a power data sampling index set, a response index set, and a frequency regulation index set of the power system are constructed to build a multi-power index set of the power system. Based on the flexible load resource adjustment scenario, each power index in the multi-power index set is classified, and the flexible load resource adjustment scenario association type of each power index is determined. Based on the preset correlation coefficient algorithm and the flexible load resource adjustment scenario, the scenario adaptation weight of each power indicator is determined; and combined with the preset power impact weight of each power indicator, the combined weight of each power indicator is determined. Based on a preset machine learning algorithm and combined with the flexible load resource adjustment scenario, the combined weight of each power index is optimized to obtain the adjustment scenario weight of each power index. Based on the adjustment scenario weight of each power index and the association type of the flexible load resource adjustment scenario, the adjustment data of the flexible load resource adjustment scenario is determined; and the flexible load resources of the power system are adjusted based on the adjustment data.

[0006] It is understood that this invention constructs a load regulation index set, a power data sampling index set, a response index set, and a frequency regulation index set using real-time operating data of the power system. This enables comprehensive modeling of the flexible load resources of the power system through multiple dimensions and types of power indicators, fully reflecting the comprehensive capabilities of flexible loads under different characteristics. Subsequently, power indicators are classified according to flexible load resource regulation scenarios. The correlation type of the flexible load resource regulation scenario distinguishes the regulation correlation of different power indicators under the current flexible load resource regulation scenario, reducing the interference of irrelevant power indicators on flexible load resource regulation. Combined weights are determined through scenario adaptation weights and power influence weights, taking into account the impact of flexible load resource regulation scenarios and power indicators on power system operation. Then, weight optimization is performed using machine learning algorithms, significantly improving the accuracy, rationality, and scenario adaptability of weight allocation. Thus, flexible load resource regulation can be carried out by adjusting scenario weights and the correlation type of flexible load resource regulation scenarios, improving the comprehensiveness and accuracy of flexible load resource regulation.

[0007] As a preferred embodiment, the real-time operating data includes: power system load data, power sampling data, power system response data, primary frequency regulation data, secondary frequency regulation data, and frequency regulation mileage data; Based on the real-time operational data, the power system's load regulation index set, power data sampling index set, response index set, and frequency regulation index set are constructed to build a multi-power index set for the power system, including: Based on the power system load data, a set of load regulation indicators for the power system is constructed; Based on the power sampling data, a power data sampling index set for the power system is constructed; Based on the power system response data and the flexible load resource adjustment scenario, the power system response index set is determined. Based on the primary frequency regulation data, secondary frequency regulation data, and frequency regulation mileage data, a set of frequency regulation indicators for the power system is constructed. Based on the load regulation index set, power data sampling index set, response index set, and frequency regulation index set, a multi-power index set for the power system is constructed.

[0008] This preferred scheme utilizes multi-dimensional data, including power system load data, power sampling data, power system response data, primary frequency regulation data, secondary frequency regulation data, and frequency regulation mileage data, to construct multi-level load regulation index sets, power data sampling index sets, response index sets, and frequency regulation index sets. This enables comprehensive modeling of the power system's flexible load resources through multiple dimensions and types of power indicators. It can comprehensively and accurately reflect the integrated capabilities of flexible loads under different characteristics, thereby improving the comprehensiveness and accuracy of flexible load resource regulation.

[0009] As a preferred embodiment, the power system load data includes: daily average load, daily maximum load, monthly average load, monthly maximum load, measured baseline load, baseline load maximum value, baseline load minimum value, electricity price data, and the operating status of load management equipment; The process of constructing a set of load regulation indicators for the power system based on the power system load data includes: The daily load factor of the power system is determined based on the ratio of the daily average load to the daily maximum load. The monthly load factor of the power system is determined based on the ratio of the monthly average load to the monthly maximum load. Based on the baseline load maximum and baseline load minimum, the daily peak-valley difference and daily peak-valley difference rate of the power system are determined; Based on the measured baseline load and electricity price data, the electricity price-load sensitivity coefficient of the power system is determined; Based on the operating status of the load management equipment, the operating score of the load management equipment of the power system is determined; Based on the daily load factor, monthly load factor, daily peak-valley difference, daily peak-valley difference rate, electricity price load sensitivity coefficient, and load management equipment operation score of the power system, the load regulation index set of the power system is determined.

[0010] This preferred scheme reflects the temporal distribution characteristics of the load through daily and monthly load factors, characterizes the load fluctuation characteristics with daily peak-valley difference and daily peak-valley difference rate, reflects the impact of the load on the electricity price with the electricity price sensitivity coefficient, and reflects the actual operating status of the load management equipment with the equipment operation score. It comprehensively and accurately depicts the load characteristics of the power system, thereby providing a data foundation for subsequent flexible load resource regulation and improving the comprehensiveness and accuracy of flexible load resource regulation.

[0011] As a preferred embodiment, the power sampling data includes: the number of power samplings, the power regulation cycle, and the power sampling data transmission time interval; The construction of the power data sampling index set of the power system based on the power sampling data includes: Based on the power sampling number and power regulation cycle, the power sampling interval of the power system is determined; and based on the power sampling interval of the power system, the power sampling frequency of the power system is determined. The power sampling data uploading frequency of the power system is determined based on the power sampling data uploading time interval; Based on the power sampling interval, power sampling frequency, and power sampling data transmission frequency of the power system, a power data sampling index set of the power system is constructed.

[0012] This preferred scheme, through power sampling interval, power sampling frequency, and power sampling data uploading frequency, can comprehensively evaluate the real-time performance and reliability of power data sampling, avoid deviations in flexible load resource regulation caused by data lag, and thus improve the comprehensiveness and accuracy of flexible load resource regulation.

[0013] As a preferred embodiment, the power system response data includes: maximum output power, minimum output power, planned output power, measured output power, and measured output power change. The determination of the power system response index set based on the power system response data and the flexible load resource adjustment scenario includes: The adjustable capacity of the power system is determined based on the maximum and minimum output power. Based on the measured baseline load, planned output value, and measured output power, and in conjunction with the flexible load resource adjustment scenario, the response capacity of the power system is determined. Based on the response capacity of the power system, determine the response rate and average response rate of the power system. Based on the measured change in output power, the ramp rate of the power system is determined; Based on the measured output power, measured baseline load, and planned output value, the response accuracy and average response accuracy of the power system are determined. Based on the measured output power and response accuracy, the response time and effective response duration of the power system are determined. Based on the adjustable capacity, response capacity, response rate, average response rate, ramp rate, response accuracy, average response accuracy, response time, and effective response duration of the power system, the set of response indicators for the power system is determined.

[0014] This preferred scheme uses adjustable capacity and response capacity to reflect the upper limit of regulation potential, response rate and ramp rate to reflect the dynamic characteristics of the regulation process, and response accuracy and response time to measure the quality of regulation. This enables the constructed set of response indicators to comprehensively and accurately evaluate the dynamic response capability of the power system under flexible load regulation scenarios, thereby improving the comprehensiveness and accuracy of flexible load resource regulation.

[0015] As a preferred embodiment, the primary frequency modulation data includes: measured frequency modulation dead zone frequency difference, maximum permissible frequency modulation frequency difference, measured frequency modulation limiting power, maximum permissible frequency modulation power, measured frequency modulation power change, measured frequency change, theoretical frequency modulation power change, theoretical frequency change, and step test response data; the secondary frequency modulation data includes: secondary frequency modulation adjustment rate, secondary frequency modulation adjustment accuracy, and secondary frequency modulation response time; The construction of the frequency regulation index set of the power system based on the primary frequency regulation data, secondary frequency regulation data, and frequency regulation mileage data includes: Based on the ratio of the measured frequency regulation dead zone frequency difference to the maximum allowable frequency regulation frequency difference, the primary frequency regulation dead zone performance index of the power system is determined. Based on the ratio of the measured frequency regulation limiting power to the maximum allowable frequency regulation power, the primary frequency regulation limiting performance index of the power system is determined; Based on the ratio of the measured frequency regulation power change to the measured frequency change, and combined with the ratio of the theoretical frequency regulation power change to the theoretical frequency change, the primary frequency regulation droop rate performance index of the power system is determined. Based on the step test response data, the primary frequency regulation dynamic performance index of the power system is determined; Based on the primary frequency regulation dead zone performance index, primary frequency regulation amplitude limiting performance index, primary frequency regulation droop rate performance index, and primary frequency regulation dynamic performance index of the power system, a primary frequency regulation index set of the power system is constructed. Based on the secondary frequency regulation rate, secondary frequency regulation accuracy, and secondary frequency regulation response time, the comprehensive frequency regulation performance index and the daily average value of the comprehensive frequency regulation performance index of the power system are determined. Based on the comprehensive frequency regulation performance index of the secondary frequency regulation of the power system and the daily average value of the comprehensive frequency regulation performance index of the secondary frequency regulation of the power system, a set of secondary frequency regulation indexes of the power system is constructed. Based on the frequency regulation mileage data, the frequency regulation mileage index of the power system is constructed; Based on the primary frequency regulation index set, secondary frequency regulation index set, and frequency regulation mileage index of the power system, a frequency regulation index set of the power system is constructed.

[0016] This preferred scheme reflects the dynamic response capability of the power system in primary frequency regulation through primary frequency regulation dead zone performance indicators, primary frequency regulation limiting performance indicators, primary frequency regulation droop rate performance indicators, and primary frequency regulation dynamic performance indicators; it reflects the frequency regulation stability and dispatchability of the power system in secondary frequency regulation through secondary frequency regulation adjustment rate, secondary frequency regulation adjustment accuracy, and secondary frequency regulation response time; and it reflects the contribution of flexible load resources in the frequency regulation dimension through the frequency regulation mileage indicator. This enables subsequent flexible load resource regulation to consider frequency control tasks with different time scales and accuracy requirements, improves the frequency stability and renewable energy absorption capacity of the power system, and thus improves the comprehensiveness and accuracy of flexible load resource regulation.

[0017] As a preferred embodiment, based on the flexible load resource adjustment scenario, each power index in the multi-power index set is classified, and the flexible load resource adjustment for each power index is determined. Based on the aforementioned flexible load resource adjustment scenario, the target value for flexible load resource adjustment of the power system is determined; Based on the preset mutual information method, calculate the correlation score between each power index in the multi-power index set and the flexible load resource adjustment target value; If the correlation score between the power index and the flexible load resource adjustment target value is greater than or equal to the preset correlation threshold, then the flexible load resource adjustment scenario association type of the power index is determined to be a core correlation index. If the correlation score between the power index and the flexible load resource adjustment target value is less than the preset correlation threshold, then the flexible load resource adjustment scenario association type of the power index is determined to be a non-core correlation index.

[0018] This preferred solution calculates the correlation score between power indicators and the target value of flexible load resource regulation using the mutual information method. By differentiating the correlation types of flexible load resource regulation scenarios, it can distinguish the regulation correlation of different power indicators under the current flexible load resource regulation scenario, reducing the interference of irrelevant power indicators on flexible load resource regulation and improving the comprehensiveness and accuracy of flexible load resource regulation.

[0019] As a preferred embodiment, the step of determining the scenario adaptation weight of each power indicator based on a preset correlation coefficient algorithm and the flexible load resource adjustment scenario; and combining the preset power impact weight of each power indicator to determine the combined weight of each power indicator, including: Construct a standardized matrix of the multiple power indicators set, and calculate the Pearson correlation coefficient between every two power indicators based on the standardized matrix. Based on the Pearson correlation coefficient between each pair of the aforementioned power indicators, a correlation coefficient matrix of the multi-power indicator set is constructed. Based on the correlation coefficient matrix, the conflict score between each power indicator and other power indicators is determined; Based on the correlation score between each of the power indicators and the flexible load resource adjustment target value, the scenario adaptation factor of each of the power indicators is determined. Based on the conflict score and scenario adaptation factor of each power indicator, the information content score of each power indicator is determined; and based on the proportion of the information content score of each power indicator to all information content scores, the scenario adaptation weight of each power indicator is determined. The scenario adaptation weight and the preset power impact weight of each power indicator are weighted and summed to determine the combined weight of each power indicator.

[0020] This preferred scheme uses conflict scores to reflect the correlation between power indicators and scenario adaptation factors to reflect the responsiveness of power indicators in flexible load resource regulation scenarios. Then, the scenario adaptation weights and power impact weights are weighted and summed to make the combined weights take into account the impact of flexible load resource regulation scenarios and power indicators on power system operation, thereby improving the comprehensiveness and accuracy of flexible load resource regulation.

[0021] As a preferred embodiment, the step of optimizing the combined weights of each power index based on a preset machine learning algorithm and the flexible load resource adjustment scenario to obtain the adjustment scenario weights of each power index includes: Obtain historical operating data of the power system in the flexible load resource adjustment scenario; Based on the preset VIKOR multi-criteria decision-making method, the historical operating data is extrapolated to determine the training labels; Based on the historical operating data and training labels, the preset LightGBM model is trained to determine the weight optimization model for flexible load resource adjustment scenarios. Based on the flexible load resource adjustment scenario weight optimization model, the combined weights of each power index are optimized to obtain the adjustment scenario weights of each power index.

[0022] This preferred solution improves the robustness of the LightGBM model training process by combining historical operating data with the VIKOR multi-criteria decision-making method. Furthermore, it optimizes the combined weights through a flexible load resource adjustment scenario weight optimization model, thereby improving the accuracy, rationality, and scenario adaptability of weight allocation and ultimately enhancing the comprehensiveness and accuracy of flexible load resource adjustment.

[0023] As a preferred embodiment, the step of determining the adjustment data for the flexible load resource adjustment scenario based on the adjustment scenario weight of each of the power indicators and the association type of the flexible load resource adjustment scenario; and adjusting the flexible load resources of the power system based on the adjustment data, includes: Based on the flexible load resource adjustment scenario association type of each of the power indicators, the total weight of the core association indicators and the total weight of the non-core association indicators of the power system are determined. If the flexible load resource adjustment scenario association type of the power indicator is a core association indicator, then the flexible load resource adjustment weight of the power indicator is determined based on the total weight of the core association indicator and the adjustment scenario weight of the power indicator. If the flexible load resource adjustment scenario association type of the power indicator is a non-core association indicator, then the flexible load resource adjustment weight of the power indicator is determined based on the total weight of the non-core association indicator and the adjustment scenario weight of the power indicator. According to the preset mapping function, the flexible load resource adjustment weight of each power index is mapped to obtain the flexible load resource adjustment mapping score of each power index. The flexible load resource adjustment mapping score and the flexible load resource adjustment weight of each power index are multiplied and summed to determine the flexible load resource adjustment potential score of the power system. Based on the flexible load resource adjustment potential score and the preset flexible load resource adjustment potential threshold, a preset database is queried to determine the adjustment data of the power system in the flexible load resource adjustment scenario; The flexible load resources of the power system are adjusted based on the adjustment data.

[0024] This preferred scheme ensures the dominant role of core related indicators in the evaluation by distinguishing the total weight of core related indicators and non-core related indicators, while also taking into account the auxiliary influence of non-core related indicators. After mapping the flexible load resource adjustment weight to the flexible load resource adjustment mapping score, the scores are multiplied and accumulated to transform the multidimensional power indicators into a single flexible load resource adjustment potential score, thereby enabling accurate and comprehensive adjustment of flexible load resources. Attached Figure Description

[0025] Figure 1 This is a flowchart illustrating the steps of a flexible load resource adjustment method based on a combination of multiple power indicators, as provided in an embodiment of the present invention. Detailed Implementation

[0026] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0027] Example 1 Please refer to Figure 1 , Figure 1 The flowchart of a flexible load resource adjustment method based on the combination and weighting of multiple power indicators provided in this embodiment of the invention includes steps S1 to S6.

[0028] Step S1: Obtain real-time operating data of the power system and flexible load resource adjustment scenarios.

[0029] Step S2: Based on the real-time operating data, construct the load regulation index set, power data sampling index set, response index set, and frequency regulation index set of the power system to construct the multi-power index set of the power system.

[0030] In this embodiment, the real-time operating data includes: power system load data, power sampling data, power system response data, primary frequency regulation data, secondary frequency regulation data, and frequency regulation mileage data; Based on the real-time operational data, the power system's load regulation index set, power data sampling index set, response index set, and frequency regulation index set are constructed to build a multi-power index set for the power system, including: Based on the power system load data, a set of load regulation indicators for the power system is constructed; Based on the power sampling data, a power data sampling index set for the power system is constructed; Based on the power system response data and the flexible load resource adjustment scenario, the power system response index set is determined. Based on the primary frequency regulation data, secondary frequency regulation data, and frequency regulation mileage data, a set of frequency regulation indicators for the power system is constructed. Based on the load regulation index set, power data sampling index set, response index set, and frequency regulation index set, a multi-power index set for the power system is constructed.

[0031] This embodiment utilizes multi-dimensional data, including power system load data, power sampling data, power system response data, primary frequency regulation data, secondary frequency regulation data, and frequency regulation mileage data, to construct multi-level load regulation index sets, power data sampling index sets, response index sets, and frequency regulation index sets. This enables comprehensive modeling of the power system's flexible load resources through multiple dimensions and types of power indicators. It can comprehensively and accurately reflect the integrated capabilities of flexible loads under different characteristics, thereby improving the comprehensiveness and accuracy of flexible load resource regulation.

[0032] In this embodiment, the power system load data includes: daily average load, daily maximum load, monthly average load, monthly maximum load, baseline load maximum, baseline load minimum, electricity price data, and the operating status of load management equipment; The process of constructing a set of load regulation indicators for the power system based on the power system load data includes: The daily load factor of the power system is determined based on the ratio of the daily average load to the daily maximum load. The monthly load factor of the power system is determined based on the ratio of the monthly average load to the monthly maximum load. Based on the baseline load maximum and baseline load minimum, the daily peak-valley difference and daily peak-valley difference rate of the power system are determined; Based on the baseline load maximum, baseline load minimum, and electricity price data, the electricity price load sensitivity coefficient of the power system is determined. Based on the operating status of the load management equipment, the operating score of the load management equipment of the power system is determined; Based on the daily load factor, monthly load factor, daily peak-valley difference, daily peak-valley difference rate, electricity price load sensitivity coefficient, and load management equipment operation score of the power system, the load regulation index set of the power system is determined.

[0033] In an optional embodiment, the power system load data includes: daily average load. Maximum daily load Monthly average load Monthly maximum load Measured baseline load, maximum baseline load Baseline load minimum Electricity price data The operating status of load management equipment includes: the remote signaling capability, telemetry capability, remote control capability, and remote adjustment capability of the load management device, which are respectively expressed as: , , and ;when , , or A value of 1 indicates that the load management device has the corresponding capability; a value of 0 indicates that it does not have the corresponding capability. , , and The weighting coefficients for remote signaling capability, telemetry capability, remote control capability, and remote adjustment capability are respectively, satisfying... ; Therefore, daily load factor for ; Monthly load factor for ; Daily peak-valley difference for ; Daily peak-valley difference rate for ; Electricity price load sensitivity coefficient for ;in, This represents the measured baseline load after normalization. This represents the normalized electricity price data; This represents the covariance of the measured baseline load and electricity price data after normalization. This represents the variance of the measured baseline load after normalization. This represents the variance of the electricity price data after normalization. Load management equipment operating score for .

[0034] This embodiment reflects the temporal distribution characteristics of the load through daily and monthly load factors, characterizes the load fluctuation characteristics with daily peak-valley difference and daily peak-valley difference rate, reflects the impact of the load on the electricity price with the electricity price sensitivity coefficient, and reflects the actual operating status of the load management equipment with the equipment operation score. It comprehensively and accurately depicts the load characteristics of the power system, thereby providing a data foundation for subsequent flexible load resource regulation and improving the comprehensiveness and accuracy of flexible load resource regulation.

[0035] In this embodiment, the power sampling data includes: the number of power samplings, the power regulation cycle, and the power sampling data transmission time interval; The construction of the power data sampling index set of the power system based on the power sampling data includes: Based on the power sampling number and power regulation cycle, the power sampling interval of the power system is determined; and based on the power sampling interval of the power system, the power sampling frequency of the power system is determined. The power sampling data uploading frequency of the power system is determined based on the power sampling data uploading time interval; Based on the power sampling interval, power sampling frequency, and power sampling data transmission frequency of the power system, a power data sampling index set of the power system is constructed.

[0036] In an optional embodiment, the power sampling data includes: the number of power samplings. Power regulation cycle and the time interval for uploading power sampling data ; Power sampling interval for ; Power sampling frequency for ; Frequency of power sampling data transmission for .

[0037] This embodiment, by adjusting the power sampling interval, power sampling frequency, and power sampling data transmission frequency, can comprehensively evaluate the real-time performance and reliability of power data sampling, avoiding deviations in flexible load resource regulation caused by data lag, thereby improving the comprehensiveness and accuracy of flexible load resource regulation.

[0038] In this embodiment, the power system response data includes: maximum output power, minimum output power, planned output power, measured output power, and measured output power change. The determination of the power system response index set based on the power system response data and the flexible load resource adjustment scenario includes: The adjustable capacity of the power system is determined based on the maximum and minimum output power. Based on the measured baseline load, planned output value, and measured output power, and in conjunction with the flexible load resource adjustment scenario, the response capacity of the power system is determined. Based on the response capacity of the power system, determine the response rate and average response rate of the power system. Based on the measured change in output power, the ramp rate of the power system is determined; Based on the measured output power, measured baseline load, and planned output value, the response accuracy and average response accuracy of the power system are determined. Based on the measured output power and response accuracy, the response time and effective response duration of the power system are determined. Based on the adjustable capacity, response capacity, response rate, average response rate, ramp rate, response accuracy, average response accuracy, response time, and effective response duration of the power system, the set of response indicators for the power system is determined.

[0039] In an optional embodiment, the power system response data includes: maximum output power. Minimum output power Planned output value Actual output power and measured output power change ; Adjustable capacity for ;in, Indicates time; There are two methods for calculating response capacity. When the flexible load resource adjustment scenario involves day-ahead and intraday hourly interactions, and it targets demand-side resources requiring a baseline load, the following method is used: Calculate response capacity When flexible load resource adjustment scenarios are medium- to long-term, minute- to second-level interactive scenarios where it is difficult to find a baseline, and when targeting demand-side resources such as load-type and energy storage / generation-type resources, the following should be used: Calculate response capacity ; Indicates in Measured baseline load at time; Response rate for ; Average response rate for ; Indicates time The specific value of the response weighting factor is determined by technical personnel based on actual needs. Climbing speed for ; Response accuracy There are two ways to calculate it, one of which is... Secondly Average response accuracy There are two ways to calculate it, one of which is... Secondly ; Response time Represented as: ;in, This refers to the moment when an action prompt or instruction requires the load management device to take action; The moment when the measured output power first reaches the specified threshold; and Representing time respectively and The measured output power; The output power coefficient represents the specified threshold value for the measured output power. Effective response time Represented as Effective response time This refers to the effective response time of demand-side resources. The parameter is between 0 and 1, representing the threshold for response accuracy; Used for statistical time response accuracy Has it been achieved? If the condition is met, the response is considered valid and is assigned a value of 1; otherwise, it is assigned a value of 0.

[0040] This embodiment uses adjustable capacity and response capacity to reflect the upper limit of regulation potential, response rate and ramp rate to reflect the dynamic characteristics of the regulation process, and response accuracy and response time to measure the quality of regulation. As a result, the constructed set of response indicators can comprehensively and accurately evaluate the dynamic response capability of the power system in flexible load regulation scenarios, thereby improving the comprehensiveness and accuracy of flexible load resource regulation.

[0041] In this embodiment, the primary frequency modulation data includes: measured frequency modulation dead zone frequency difference, maximum permissible frequency modulation frequency difference, measured frequency modulation limiting power, maximum permissible frequency modulation power, measured frequency modulation power change, measured frequency change, theoretical frequency modulation power change, theoretical frequency change, and step test response data; the secondary frequency modulation data includes: secondary frequency modulation adjustment rate, secondary frequency modulation adjustment accuracy, and secondary frequency modulation response time. The construction of the frequency regulation index set of the power system based on the primary frequency regulation data, secondary frequency regulation data, and frequency regulation mileage data includes: Based on the ratio of the measured frequency regulation dead zone frequency difference to the maximum allowable frequency regulation frequency difference, the primary frequency regulation dead zone performance index of the power system is determined. Based on the ratio of the measured frequency regulation limiting power to the maximum allowable frequency regulation power, the primary frequency regulation limiting performance index of the power system is determined; Based on the ratio of the measured frequency regulation power change to the measured frequency change, and combined with the ratio of the theoretical frequency regulation power change to the theoretical frequency change, the primary frequency regulation droop rate performance index of the power system is determined. Based on the step test response data, the primary frequency regulation dynamic performance index of the power system is determined; Based on the primary frequency regulation dead zone performance index, primary frequency regulation amplitude limiting performance index, primary frequency regulation droop rate performance index, and primary frequency regulation dynamic performance index of the power system, a primary frequency regulation index set of the power system is constructed. Based on the secondary frequency regulation rate, secondary frequency regulation accuracy, and secondary frequency regulation response time, the comprehensive frequency regulation performance index and the daily average value of the comprehensive frequency regulation performance index of the power system are determined. Based on the comprehensive frequency regulation performance index of the secondary frequency regulation of the power system and the daily average value of the comprehensive frequency regulation performance index of the secondary frequency regulation of the power system, a set of secondary frequency regulation indexes of the power system is constructed. Based on the frequency regulation mileage data, the frequency regulation mileage index of the power system is constructed; Based on the primary frequency regulation index set, secondary frequency regulation index set, and frequency regulation mileage index of the power system, a frequency regulation index set of the power system is constructed.

[0042] In an optional embodiment, the primary frequency modulation data includes: measured frequency modulation dead zone frequency difference. Maximum permissible frequency difference Measured frequency modulation limiting power Maximum permissible frequency modulation power Measured frequency modulation power variation Measured frequency change Theoretical frequency modulation power variation Theoretical frequency change and step test response data; the secondary frequency modulation data includes: secondary frequency modulation adjustment rate. Secondary frequency modulation adjustment accuracy and secondary frequency modulation response time ; Set here Represents the first power system One frequency modulation unit; Represents the first power system Secondary adjustment; Therefore, the performance index of primary frequency modulation dead time Represented as ; Primary frequency modulation limiting performance indicators Represented as ; Primary frequency modulation droop rate performance index Represented as ; The dynamic performance of a single frequency modulation refers to the dynamic response characteristics of the controlled variable in a step test, including rise time, lag time, settling time, and active power regulation deviation. Primary frequency modulation dynamic performance indicators Represented as ;in, The rise time is a performance metric. ; Ascending time; Maximum allowable rise time; The performance indicator is the lag time; ; The time lag is the time delay. This is the maximum permissible delay time; To adjust the performance indicators of time; ; To adjust the time; Indicates the maximum allowable adjustment time; This is a performance indicator for active power regulation deviation. ; Indicates active power regulation deviation; Indicates the maximum permissible adjustment deviation; Secondary frequency modulation comprehensive frequency modulation performance indicators Represented as ; Daily average value of integrated frequency modulation performance index for secondary frequency modulation Represented as ; To adjust the number of times; Secondary frequency modulation adjustment rate This refers to the rate at which the load demand response to the setpoint command is measured by the frequency modulation unit. No. The degree to which the actual adjustment speed in response to the setpoint command during the adjustment process reaches its standard speed compared to the speed it should have reached is calculated as follows: ; Secondary frequency modulation adjustment accuracy This refers to the difference between the actual output and the setpoint output after the load demand response has stabilized; it measures the frequency regulation unit. No. The degree to which the actual adjustment deviation is achieved during the adjustment process compared to its allowable deviation is calculated as follows: ; Second frequency modulation response time This refers to the time required for the load demand output to reliably cross the regulation dead zone in the same direction as the original output point after the system issues a command. It measures the frequency modulation unit. No. The degree to which the actual response time relative to the standard response time is achieved during the adjustment process is calculated as follows: ; FM mileage indicators Represented as ;in, Indicates the frequency modulation unit number The frequency modulation mileage of each adjustment.

[0043] This embodiment reflects the dynamic response capability of the power system in primary frequency regulation through primary frequency regulation dead zone performance indicators, primary frequency regulation limiting performance indicators, primary frequency regulation droop rate performance indicators, and primary frequency regulation dynamic performance indicators; it reflects the frequency regulation stability and dispatchability of the power system in secondary frequency regulation through secondary frequency regulation adjustment rate, secondary frequency regulation adjustment accuracy, and secondary frequency regulation response time; and it reflects the contribution of flexible load resources in the frequency regulation dimension through the frequency regulation mileage indicator. This enables subsequent flexible load resource regulation to consider frequency control tasks with different time scales and accuracy requirements, improves the frequency stability and renewable energy absorption capacity of the power system, and thus improves the comprehensiveness and accuracy of flexible load resource regulation.

[0044] Step S3: Based on the flexible load resource adjustment scenario, classify each power index in the multi-power index set and determine the flexible load resource adjustment scenario association type for each power index.

[0045] In this embodiment, based on the flexible load resource adjustment scenario, each power index in the multi-power index set is classified to determine the flexible load resource adjustment for each power index. Based on the aforementioned flexible load resource adjustment scenario, the target value for flexible load resource adjustment of the power system is determined; Based on the preset mutual information method, calculate the correlation score between each power index in the multi-power index set and the flexible load resource adjustment target value; If the correlation score between the power index and the flexible load resource adjustment target value is greater than or equal to the preset correlation threshold, then the flexible load resource adjustment scenario association type of the power index is determined to be a core correlation index. If the correlation score between the power index and the flexible load resource adjustment target value is less than the preset correlation threshold, then the flexible load resource adjustment scenario association type of the power index is determined to be a non-core correlation index.

[0046] It should be noted that the mutual information method is a statistical method used to quantify the degree of interdependence between two random variables; it measures the amount of information one variable contains about another, or in other words, the degree to which knowing the value of one variable reduces the uncertainty about the other. A higher mutual information value indicates a stronger dependency between the two variables.

[0047] In one optional embodiment, different flexible load resource adjustment scenarios have different flexible load resource adjustment target values. For example, in the peak shaving scenario, the flexible load resource adjustment target value can be quantified as the "average peak-to-valley difference reduction rate"; in the frequency regulation scenario, the flexible load resource adjustment target value can be quantified as the "root mean square value of frequency deviation". After determining the target value for flexible load resource regulation, the correlation score between power indicators and the target value for flexible load resource regulation is calculated using the mutual information method; specifically, the target value for flexible load resource regulation is set as follows: , Indicates the first Each electricity indicator; the correlation score of the indicator is calculated using the mutual information method. ,Right now Set the indicator correlation threshold to 0.5; if the indicator correlation score is... If the correlation threshold is greater than or equal to the indicator's threshold, then the electricity indicator is considered valid. The flexible load resource adjustment scenario correlation type is the core correlation indicator; otherwise, it is considered a non-core correlation indicator.

[0048] This embodiment calculates the correlation score between power indicators and the target value of flexible load resource adjustment using the mutual information method. By differentiating the correlation types of flexible load resource adjustment scenarios, it can distinguish the adjustment correlation of different power indicators under the current flexible load resource adjustment scenario, reduce the interference of irrelevant power indicators on flexible load resource adjustment, and improve the comprehensiveness and accuracy of flexible load resource adjustment.

[0049] Step S4: Based on the preset correlation coefficient algorithm and the flexible load resource adjustment scenario, determine the scenario adaptation weight of each power indicator; and combine the preset power influence weight of each power indicator to determine the combined weight of each power indicator.

[0050] In this embodiment, the step of determining the scenario adaptation weight of each power indicator based on a preset correlation coefficient algorithm and the flexible load resource adjustment scenario, and combining the preset power impact weight of each power indicator to determine the combined weight of each power indicator, includes: Construct a standardized matrix of the multiple power indicators set, and calculate the Pearson correlation coefficient between every two power indicators based on the standardized matrix. Based on the Pearson correlation coefficient between each pair of the aforementioned power indicators, a correlation coefficient matrix of the multi-power indicator set is constructed. Based on the correlation coefficient matrix, the conflict score between each power indicator and other power indicators is determined; Based on the correlation score between each of the power indicators and the flexible load resource adjustment target value, the scenario adaptation factor of each of the power indicators is determined. Based on the conflict score and scenario adaptation factor of each power indicator, the information content score of each power indicator is determined; and based on the proportion of the information content score of each power indicator to all information content scores, the scenario adaptation weight of each power indicator is determined. The scenario adaptation weight and the preset power impact weight of each power indicator are weighted and summed to determine the combined weight of each power indicator.

[0051] In an optional embodiment, a commonality is set There are several electricity indicators, and each electricity indicator has... Each sample of observations allows for the construction of the original power index data matrix. ;in, Indicates the first The first electricity index Individual sample observations; ; For each electricity index Perform z-score standardization: ; (Electricity indicators) (mean of the sample observations) (Electricity indicators) (the standard deviation of the sample observations); therefore, the standardized matrix of the power index is represented as: ; For any two electricity indicators and Calculate the Pearson correlation coefficient between the two electricity indicators, i.e. Due to z-score standardization, With a variance of 1, the Pearson correlation coefficient between the two electricity indicators can be simplified to: ; Therefore, it can form Correlation coefficient matrix: Among them, diagonal elements (That is, the electricity index is completely correlated with itself). Conflict score can be represented as: ; The Pearson correlation coefficient between the two electricity indicators; The larger the value, the higher the power index. The stronger the conflict with other electricity indicators; The scene adaptation factor is expressed as: ;in, This represents a scenario for flexible load resource adjustment. Therefore, the information content score is expressed as: ; The scene adaptation weight is represented as follows: ; Subsequently, based on expert evaluation, the power impact weights of power indicators were determined. ; The preference coefficient is determined by optimizing the method based on expert confidence or entropy. The combined weights are expressed as: .

[0052] This embodiment uses conflict scores to reflect the correlation between power indicators and scenario adaptation factors to reflect the responsiveness of power indicators in flexible load resource adjustment scenarios. Then, the scenario adaptation weights and power impact weights are weighted and summed to make the combined weights take into account the impact of flexible load resource adjustment scenarios and power indicators on power system operation, thereby improving the comprehensiveness and accuracy of flexible load resource adjustment.

[0053] Step S5: Based on the preset machine learning algorithm and combined with the flexible load resource adjustment scenario, optimize the combined weight of each power index to obtain the adjustment scenario weight of each power index.

[0054] In this embodiment, the step of optimizing the combined weights of each power index according to a preset machine learning algorithm and in conjunction with the flexible load resource adjustment scenario to obtain the adjustment scenario weights of each power index includes: Obtain historical operating data of the power system in the flexible load resource adjustment scenario; Based on the preset VIKOR multi-criteria decision-making method, the historical operating data is extrapolated to determine the training labels; Based on the historical operating data and training labels, the preset LightGBM model is trained to determine the weight optimization model for flexible load resource adjustment scenarios. Based on the flexible load resource adjustment scenario weight optimization model, the combined weights of each power index are optimized to obtain the adjustment scenario weights of each power index.

[0055] It's worth noting that LightGBM (Light Gradient Boosting Machine) is a highly efficient gradient boosting decision tree (GBDT) algorithm framework. Through innovative techniques such as histogram-based decision tree optimization, one-sided gradient sampling (GOSS), and mutually exclusive feature binding (EFB), it significantly reduces computational resource consumption and memory usage while maintaining high accuracy. Compared to traditional GBDT implementations (such as XGBoost), LightGBM is faster and more efficient in processing large-scale data, making it particularly suitable for high-dimensional sparse data scenarios. VIKOR (VlseKriterijumska Optimizacija IKompromisno Resenje) is a multi-criteria decision analysis method that aims to find the closest solution to the ideal solution through compromise. Its core idea is to determine the positive ideal solution (optimal value of each criterion) and the negative ideal solution (worst value of each criterion). By calculating the distance between each solution and both, and combining two indicators—group utility (overall satisfaction) and individual regret (maximum dissatisfaction)—a ranking index Q-value is generated. Finally, the solution with the smallest Q-value that satisfies the stability condition is selected as the optimal compromise solution.

[0056] In one optional embodiment, historical operating data of the power system in the flexible load resource adjustment scenario is acquired, and the optimal weight vector under the flexible load resource adjustment scenario is obtained by extrapolating the historical operating data using the VIKOR multi-criteria decision method, which serves as the training label; training data is constructed using the historical operating data, and in this embodiment, scenario feature vectors are used; the LightGBM model is trained using the scenario feature vectors and training labels to determine the weight optimization model for the flexible load resource adjustment scenario; since there are already mature model training processes in the prior art, this embodiment will not elaborate further here; After obtaining the flexible load resource adjustment scenario weight optimization model, a scenario feature vector corresponding to the current flexible load resource adjustment scenario is constructed and input into the flexible load resource adjustment scenario weight optimization model to obtain the dynamic adjustment coefficient of each power index. ; Therefore, adjusting the scene weights is expressed as: .

[0057] This embodiment improves the robustness of the LightGBM model training process by combining historical operating data with the VIKOR multi-criteria decision-making method. Furthermore, it optimizes the combined weights through a flexible load resource adjustment scenario weight optimization model, thereby improving the accuracy, rationality, and scenario adaptability of weight allocation and ultimately enhancing the comprehensiveness and accuracy of flexible load resource adjustment.

[0058] Step S6: Based on the adjustment scenario weight of each power index and the association type of the flexible load resource adjustment scenario, determine the adjustment data of the flexible load resource adjustment scenario; and adjust the flexible load resources of the power system based on the adjustment data.

[0059] In this embodiment, determining the adjustment data for the flexible load resource adjustment scenario based on the adjustment scenario weight of each power index and the association type of the flexible load resource adjustment scenario, and adjusting the flexible load resources of the power system based on the adjustment data, includes: Based on the flexible load resource adjustment scenario association type of each of the power indicators, the total weight of the core association indicators and the total weight of the non-core association indicators of the power system are determined. If the flexible load resource adjustment scenario association type of the power indicator is a core association indicator, then the flexible load resource adjustment weight of the power indicator is determined based on the total weight of the core association indicator and the adjustment scenario weight of the power indicator. If the flexible load resource adjustment scenario association type of the power indicator is a non-core association indicator, then the flexible load resource adjustment weight of the power indicator is determined based on the total weight of the non-core association indicator and the adjustment scenario weight of the power indicator. According to the preset mapping function, the flexible load resource adjustment weight of each power index is mapped to obtain the flexible load resource adjustment mapping score of each power index. The flexible load resource adjustment mapping score and the flexible load resource adjustment weight of each power index are multiplied and summed to determine the flexible load resource adjustment potential score of the power system. Based on the flexible load resource adjustment potential score and the preset flexible load resource adjustment potential threshold, a preset database is queried to determine the adjustment data of the power system in the flexible load resource adjustment scenario; The flexible load resources of the power system are adjusted based on the adjustment data.

[0060] In an optional embodiment, the number of core correlation indicators for flexible load resource adjustment scenarios is calculated based on the correlation type of each power indicator. Therefore, the total weight of the core related indicators is calculated. Among them, setting the balance coefficient And set the "scene criticality"; its specific value can be set by technical personnel according to actual needs; ; Therefore, the total weight of non-core related indicators is: ; If the flexible load resource adjustment scenario correlation type of the power index is a core correlation index, its flexible load resource adjustment weight is calculated as follows: ; If the flexible load resource adjustment scenario correlation type of the power index is a non-core correlation index, its flexible load resource adjustment weight is calculated as follows: ; Then, the S-shaped utility function (whose calculation formula is usually expressed as: , The slope Mapping the flexible load resource adjustment weights of each power indicator (with the inflection point as the reference point) yields the flexible load resource adjustment mapping score for each power indicator. The flexible load resource adjustment mapping score for the core related power indicators is expressed as follows: The flexible load resource adjustment mapping score of non-core related indicators of the power sector is expressed as follows: ; Therefore, the number of electricity indicators that are not core related indicators is set as follows: The number of electricity indicators among the core related indicators is The flexible load resource adjustment potential score is expressed as: ; The threshold values ​​for flexible load resource adjustment potential are set to 0.6 and 0.8; then, the database is queried, and if... If the adjusted data is classified as Grade A, it is considered a high-quality resource and will be prioritized for use; otherwise... If the adjustment data is determined to be Grade B, indicating good resources and a need for major replenishment; if... If the data is classified as Level C, it indicates a restricted resource, which should be reserved or temporarily not used. Therefore, the regulation of the flexible load resources of the power system can be expressed as follows: if it is Level A, a high-quality resource, it should be prioritized for use, and a full regulation instruction should be issued to the flexible load resources of the power system; if it is Level B, a good resource, it should be used as a main supplement, and a regulation instruction allocated proportionally should be issued to the flexible load resources of the power system; if it is Level C, a restricted resource, it should be reserved or temporarily not used, and a regulation instruction should be issued to the flexible load resources of the power system that it should not be used or should only be used as a backup.

[0061] This embodiment ensures the leading role of core related indicators in the evaluation by distinguishing the total weight of core related indicators and the total weight of non-core related indicators, while taking into account the auxiliary influence of non-core related indicators. After mapping the flexible load resource adjustment weight to the flexible load resource adjustment mapping score, the scores are multiplied and accumulated to transform the multidimensional power indicators into a single flexible load resource adjustment potential score, thereby enabling accurate and comprehensive adjustment of flexible load resources.

[0062] This embodiment constructs a load regulation index set, a power data sampling index set, a response index set, and a frequency regulation index set using real-time operation data of the power system. This enables comprehensive modeling of the power system's flexible load resources through multiple dimensions and types of power indicators, fully reflecting the comprehensive capabilities of flexible loads under different characteristics. Subsequently, power indicators are classified according to flexible load resource regulation scenarios. The correlation type of the flexible load resource regulation scenario distinguishes the regulation correlation of different power indicators under the current flexible load resource regulation scenario, reducing the interference of irrelevant power indicators on flexible load resource regulation. Combined weights are determined by scenario adaptation weights and power influence weights, taking into account the impact of flexible load resource regulation scenarios and power indicators on power system operation. Then, weight optimization is performed using machine learning algorithms, significantly improving the accuracy, rationality, and scenario adaptability of weight allocation. Therefore, flexible load resource regulation can be carried out by adjusting scenario weights and the correlation type of flexible load resource regulation scenarios, improving the comprehensiveness and accuracy of flexible load resource regulation.

[0063] Example 2 In the flexible load resource adjustment method based on the combination and weighting of multiple power indicators provided in this embodiment, this embodiment provides a specific practical scenario for further explanation.

[0064] For a 10MW / 20MWh energy storage system in an industrial park operating during the summer daytime peak shaving period (12:00-14:00) of the regional power grid, some indicators from the power indicators in Example 1 are selected for illustrative purposes. The core related indicators are assumed to be: adjustable capacity, response rate, and effective response time; the non-core related indicators are: power sampling frequency, load management equipment operating score, response accuracy, and power sampling interval. Table 1 shows the measured data for the core and non-core related indicators. Table 1 Furthermore, the data in Table 1 can be normalized to obtain the normalized measured data of the core correlation indicators and non-core correlation indicators as shown in Table 2. Table 2 Subsequently, based on the flexible load resource adjustment method based on the combination weighting of multiple power indicators provided in the embodiment, the core related indicators and non-core related indicators shown in Table 3 are calculated, and their respective combination weights, adjustment scenario weights, and flexible load resource adjustment weights are calculated. Table 3 Subsequently, based on the flexible load resource adjustment weight, the flexible load resource adjustment potential score of the power system is calculated to be 0.8475; therefore, the adjustment data is classified as Grade A, a high-quality resource, and should be prioritized for use. At this point, a full adjustment instruction for the flexible load resource is issued.

[0065] In summary, this invention constructs a load regulation index set, a power data sampling index set, a response index set, and a frequency regulation index set using real-time operating data of the power system. This enables comprehensive modeling of the flexible load resources of the power system through multiple dimensions and types of power indicators, fully reflecting the comprehensive capabilities of flexible loads under different characteristics. Subsequently, power indicators are classified according to flexible load resource regulation scenarios. The correlation type of the flexible load resource regulation scenario distinguishes the regulation correlation of different power indicators under the current flexible load resource regulation scenario, reducing the interference of irrelevant power indicators on flexible load resource regulation. Combined weights are determined by scenario adaptation weights and power influence weights, taking into account the impact of flexible load resource regulation scenarios and power indicators on power system operation. Then, weight optimization is performed using machine learning algorithms, significantly improving the accuracy, rationality, and scenario adaptability of weight allocation. Therefore, flexible load resource regulation can be carried out by adjusting scenario weights and the correlation type of flexible load resource regulation scenarios, improving the comprehensiveness and accuracy of flexible load resource regulation.

[0066] The specific embodiments described above further illustrate the purpose, technical solution, and beneficial effects of the present invention. It should be understood that the above descriptions are merely specific embodiments of the present invention and are not intended to limit the scope of protection of the present invention. In particular, it should be noted that any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention for those skilled in the art.

Claims

1. A flexible load resource regulation method based on a multi-power index combination weighting, characterized in that, The method comprises the following steps: acquiring real-time operation data of a power system and a flexible load resource regulation scenario; based on the real-time operation data, constructing a load regulation index set, a power data sampling index set, a response index set and a frequency regulation index set of the power system to construct a multi-power index set of the power system; based on the flexible load resource regulation scenario, classifying each power index in the multi-power index set to determine the flexible load resource regulation scenario association type of each power index; based on a preset correlation coefficient algorithm and the flexible load resource regulation scenario, determining the scenario adaptation weight of each power index; combined with the preset power influence weight of each power index, determining the combined weight of each power index; according to a preset machine learning algorithm, combining the flexible load resource regulation scenario to optimize the combined weight of each power index to obtain the regulation scenario weight of each power index; based on the regulation scenario weight of each power index and the flexible load resource regulation scenario association type, determining the regulation data of the flexible load resource regulation scenario; and based on the regulation data, regulating the flexible load resource of the power system. 2.The flexible load resource regulation method based on the combination weighting of multiple power indexes according to claim 1, wherein, The real-time operation data includes power system load data, power sampling data, power system response data, primary frequency regulation data, secondary frequency regulation data and frequency regulation mileage data; based on the real-time operation data, constructing a load regulation index set, a power data sampling index set, a response index set and a frequency regulation index set of the power system to construct a multi-power index set of the power system, comprising: based on the power system load data, constructing a load regulation index set of the power system; based on the power sampling data, constructing a power data sampling index set of the power system; based on the power system response data, combining the flexible load resource regulation scenario to determine the response index set of the power system; based on the primary frequency regulation data, secondary frequency regulation data and frequency regulation mileage data, constructing a frequency regulation index set of the power system; based on the load regulation index set, power data sampling index set, response index set and frequency regulation index set, constructing a multi-power index set of the power system. 3.The flexible load resource regulation method based on the combination weighting of multiple power indexes according to claim 2, wherein, The power system load data includes daily average load, daily maximum load, monthly average load, monthly maximum load, measured baseline load, baseline load maximum value, baseline load minimum value, electricity price data and load management equipment operating state; based on the power system load data, constructing a load regulation index set of the power system, comprising: based on the ratio of the daily average load and the daily maximum load, determining the daily load coefficient of the power system; based on the ratio of the monthly average load and the monthly maximum load, determining the monthly load coefficient of the power system; based on the baseline load maximum value and the baseline load minimum value, determining the daily peak valley difference and the daily peak valley difference rate of the power system; based on the measured baseline load and the electricity price data, determining the electricity price load sensitivity coefficient of the power system; based on the load management equipment operating state, determining the load management equipment operating score of the power system; Determine a load adjustment index set of the power system based on a daily load factor, a monthly load factor, a daily peak-valley difference, a daily peak-valley difference rate, a price load sensitivity coefficient, and a load management device operation score of the power system. 4.The flexible load resource regulation method based on the combination weighting of multiple power indexes according to claim 2, wherein, The power sampling data comprises: a power sampling frequency, a power adjustment period, and a power sampling data uploading time interval; The power sampling data comprises: a power sampling frequency, a power adjustment period, and a power sampling data uploading time interval; Determine a power sampling interval of the power system based on the power sampling frequency and the power adjustment period, and determine a power sampling frequency of the power system based on the power sampling interval of the power system; Determine a power sampling data uploading frequency of the power system based on the power sampling data uploading time interval; Determine a power data sampling index set of the power system based on the power sampling interval, the power sampling frequency, and the power sampling data uploading frequency of the power system.

5. The flexible load resource regulation method based on the combination weighting of multiple power indicators according to claim 3, characterized in that, The power system response data comprises: a maximum output power, a minimum output power, a planned output value, a measured output power, and a measured output power change amount; Determine a response index set of the power system based on the power system response data and in combination with the flexible load resource adjustment scenario, comprising: Determine an adjustable capacity of the power system based on the maximum output power and the minimum output power; Determine a response capacity of the power system based on the measured baseline load, the planned output value, and the measured output power in combination with the flexible load resource adjustment scenario; Determine a response rate and an average response rate of the power system based on the response capacity of the power system; Determine a ramping rate of the power system based on the measured output power change amount; Determine a response accuracy and an average response accuracy of the power system based on the measured output power, the measured baseline load, and the planned output value; Determine a response time and an effective response duration of the power system based on the measured output power and the response accuracy; Determine a response index set of the power system based on the adjustable capacity, the response capacity, the response rate, the average response rate, the ramping rate, the response accuracy, the average response accuracy, the response time, and the effective response duration of the power system.

6. The flexible load resource regulation method based on the combination weighting of multiple power indexes according to claim 2, wherein, The primary frequency modulation data comprises: a measured frequency modulation dead zone frequency difference, a maximum allowed frequency modulation frequency difference, a measured frequency modulation amplitude limiting power, a maximum allowed frequency modulation power, a measured frequency modulation power change amount, a measured frequency change amount, a theoretical frequency modulation power change amount, a theoretical frequency change amount, and step test response data; and the secondary frequency modulation data comprises: a secondary frequency modulation adjustment rate, a secondary frequency modulation adjustment accuracy, and a secondary frequency modulation response time; Determine a frequency modulation index set of the power system based on the primary frequency modulation data, the secondary frequency modulation data, and the frequency modulation mileage data, comprising: Determine a primary frequency modulation dead zone performance index of the power system based on a ratio of the measured frequency modulation dead zone frequency difference and the maximum allowed frequency modulation frequency difference; Determine a primary frequency modulation amplitude limiting performance index of the power system based on a ratio of the measured frequency modulation amplitude limiting power and the maximum allowed frequency modulation power; Determine the primary frequency modulation adjustment rate performance index of the power system based on the ratio of the measured frequency modulation power change and the measured frequency change, and the ratio of the theoretical frequency modulation power change and the theoretical frequency change; Determine the primary frequency modulation dynamic performance index of the power system based on the step test response data; Based on the primary frequency modulation dead zone performance index, the primary frequency modulation limit performance index, the primary frequency modulation adjustment rate performance index and the primary frequency modulation dynamic performance index of the power system, construct the primary frequency modulation index set of the power system; Determine the secondary frequency modulation comprehensive frequency modulation performance index and the daily average of the secondary frequency modulation comprehensive frequency modulation performance index of the power system based on the secondary frequency modulation adjustment rate, the secondary frequency modulation adjustment accuracy and the secondary frequency modulation response time; Based on the secondary frequency modulation comprehensive frequency modulation performance index and the daily average of the secondary frequency modulation comprehensive frequency modulation performance index of the power system, construct the secondary frequency modulation index set of the power system; Based on the frequency modulation mileage data, construct the frequency modulation mileage index of the power system; Based on the primary frequency modulation index set, the secondary frequency modulation index set and the frequency modulation mileage index of the power system, construct the frequency modulation index set of the power system.

7. The flexible load resource regulation method based on the combination weighting of multiple power indexes according to claim 1, wherein, Based on the flexible load resource adjustment scene, classify each power index in the multiple power index set, and determine the flexible load resource adjustment scene association of each power index, including: Determine the flexible load resource adjustment target value of the power system based on the flexible load resource adjustment scene; According to the preset mutual information method, calculate the index correlation score between each power index in the multiple power index set and the flexible load resource adjustment target value; If the index correlation score between the power index and the flexible load resource adjustment target value is greater than or equal to the preset index correlation threshold, it is determined that the flexible load resource adjustment scene association type of the power index is a core associated index; If the index correlation score between the power index and the flexible load resource adjustment target value is less than the preset index correlation threshold, it is determined that the flexible load resource adjustment scene association type of the power index is a non-core associated index. 8.The flexible load resource regulation method based on the combination weighting of multiple power indexes according to claim 7, wherein, Determine the scene adaptation weight of each power index based on the preset correlation coefficient algorithm and the flexible load resource adjustment scene; And combined with the preset power influence weight of each power index, determine the combined weight of each power index, including: Construct the power index standardization matrix of the multiple power index set, and based on the power index standardization matrix, calculate the Pearson correlation coefficient between each two power indexes; Based on the Pearson correlation coefficient between each two power indexes, construct the correlation coefficient matrix of the multiple power index set; Determine the conflict score of each power index and other power indexes based on the correlation coefficient matrix; Determine the scene adaptation factor of each power index based on the index correlation score between each power index and the flexible load resource adjustment target value; determine an information quantity score of each power index based on the conflict score and the scenario adaptation factor of each power index; and determine a scenario adaptation weight of each power index based on a proportion of the information quantity score of each power index in all information quantity scores; perform weighted summation on the scenario adaptation weight and a preset power influence weight of each power index to determine a combined weight of each power index. 9.The flexible load resource regulation method based on the combination weighting of multiple power indexes according to claim 8, wherein, The combined weight of each power index is optimized according to a preset machine learning algorithm in combination with the flexible load resource adjustment scenario to obtain an adjustment scenario weight of each power index, including: obtaining historical operation data of the power system in the flexible load resource adjustment scenario; based on a preset VIKOR multi-criteria decision method, the historical operation data is calculated to determine the training label; based on the historical operation data and the training label, a preset LightGBM model is trained to determine a flexible load resource adjustment scenario weight optimization model; based on the flexible load resource adjustment scenario weight optimization model, the combined weight of each power index is optimized to obtain the adjustment scenario weight of each power index. 10.The flexible load resource regulation method based on the combination weighting of multiple power indexes according to claim 7, wherein, The adjustment data of the flexible load resource adjustment scenario is determined based on the adjustment scenario weight of each power index and the flexible load resource adjustment scenario association type; and based on the adjustment data, the flexible load resource of the power system is adjusted, including: based on the flexible load resource adjustment scenario association type of each power index, the total weight of the core associated index and the total weight of the non-core associated index of the power system are determined; if the flexible load resource adjustment scenario association type of the power index is a core associated index, the flexible load resource adjustment weight of the power index is determined based on the total weight of the core associated index in combination with the adjustment scenario weight of the power index; if the flexible load resource adjustment scenario association type of the power index is a non-core associated index, the flexible load resource adjustment weight of the power index is determined based on the total weight of the non-core associated index in combination with the adjustment scenario weight of the power index; According to a preset mapping function, the flexible load resource adjustment weight of each power index is mapped to obtain a flexible load resource adjustment mapping score of each power index. The flexible load resource adjustment mapping score and the flexible load resource adjustment weight of each power index are multiplied and then accumulated to determine a flexible load resource adjustment potential score of the power system. Based on the flexible load resource adjustment potential score and a preset flexible load resource adjustment potential threshold, a preset database is queried to determine the adjustment data of the power system in the flexible load resource adjustment scenario; based on the adjustment data, the flexible load resource of the power system is adjusted.

Images