A Distributed Energy Adjustability Prediction Method Based on Multimodal Data Fusion

Abstract

This invention belongs to the field of distributed energy dispatch and prediction technology, and discloses a method for predicting the adjustable capacity of distributed energy based on multimodal data fusion. The method includes collecting multimodal raw datasets, extracting and correcting spatiotemporal heterogeneous attenuation characteristic parameters, extracting grid micro-oscillation characteristic parameters, calculating the thermal accumulation attenuation coefficient and coupling coefficient, and substituting them into the prediction formula to obtain the adjustable active and reactive power capabilities. This invention solves the problem that existing technologies do not consider the linkage between spatiotemporal attenuation, thermal accumulation, and grid micro-oscillations, achieving accurate prediction through core innovative technical points and ensuring the reliability of distributed energy dispatch.

Description

Technical Field

[0001] This invention belongs to the field of distributed energy dispatch and prediction technology, specifically relating to a method for predicting the adjustable capacity of distributed energy based on multimodal data fusion. Background Technology

[0002] The core problem with existing distributed energy resource adjustability prediction technologies is the failure to consider the interplay between spatiotemporal heterogeneous attenuation of multimodal data and the combined effects of equipment thermal accumulation and grid micro-oscillations, leading to inherent flaws in the prediction schemes. Current technologies only perform simple fusion predictions based on electrical parameters and conventional environmental parameters, without addressing the spatiotemporal heterogeneous attenuation during data acquisition and transmission. They ignore raw data deviations caused by factors such as distance between acquisition nodes and transmission delays. Furthermore, they calculate the impact of equipment thermal accumulation or grid micro-oscillations separately without establishing a quantitative model linking these two factors. This fails to accurately characterize the progressive effects of multiple factors on adjustability, resulting in incomplete prediction logic and an inability to adapt to complex operating scenarios.

[0003] Based on the above problems, there is an urgent need for a technical solution that can resolve this linkage effect and improve the prediction logic. Summary of the Invention

[0004] The purpose of this invention is to address the shortcomings of existing technologies by proposing a method for predicting the adjustable capability of distributed energy resources based on multimodal data fusion. The method includes the following steps: collecting multimodal raw datasets of distributed energy resources through multiple sensors, metering devices, and power grid monitoring terminals. These raw datasets include electrical parameters, equipment thermal parameters, power grid oscillation parameters, and spatiotemporal acquisition parameters. The spatiotemporal heterogeneous attenuation characteristic parameters of the raw datasets are extracted. These parameters include spatial attenuation factors, temporal attenuation factors, data type heterogeneity weights, and sampling frequency heterogeneity coefficients. The raw datasets are then corrected using a multimodal data spatiotemporal heterogeneous attenuation correction formula to obtain the corrected dataset. The system first establishes a basic operating parameter set; then, it performs frequency domain decomposition on the grid oscillation parameters in the corrected basic operating parameter set to extract grid micro-oscillation characteristic parameters, including micro-oscillation amplitude, micro-oscillation angular frequency, and micro-oscillation attenuation coefficient; finally, it substitutes the corrected basic operating parameter set and the grid micro-oscillation characteristic parameters into the equipment thermal accumulation attenuation coefficient calculation formula to obtain the thermal accumulation attenuation coefficient; based on the thermal accumulation attenuation coefficient and the micro-oscillation attenuation coefficient, it calculates the thermal accumulation oscillation coupling coefficient; finally, it substitutes the corrected basic operating parameter set, the thermal accumulation attenuation coefficient, and the thermal accumulation oscillation coupling coefficient into the coupled adjustable capability comprehensive prediction formula to obtain the active power adjustable capability and reactive power adjustable capability of distributed energy.

[0005] Preferably, the electrical parameters include active power, reactive power, and operating current; the equipment thermal parameters include device heat flux density, actual device temperature, and ambient temperature; the grid oscillation parameters include oscillation amplitude and oscillation frequency; and the spatiotemporal acquisition parameters include acquisition node distance, data transmission delay, and sampling frequency. The acquisition node distance is the physical distance between the distributed energy data acquisition node and the dispatch terminal; the data transmission delay is the time it takes for multimodal raw data to be transmitted from the acquisition node to the dispatch terminal; and the sampling frequency is the acquisition frequency of the multimodal raw data.

[0006] More preferably, the corrected basic operating parameter set includes corrected electrical parameters, corrected equipment thermal parameters, and corrected grid oscillation parameters. The corrected electrical parameters are obtained by calculating the electrical parameters using a multimodal data spatiotemporal heterogeneous attenuation correction formula. The corrected equipment thermal parameters are obtained by calculating the equipment thermal parameters using a multimodal data spatiotemporal heterogeneous attenuation correction formula. The corrected grid oscillation parameters are obtained by calculating the grid oscillation parameters using a multimodal data spatiotemporal heterogeneous attenuation correction formula.

[0007] More preferably, the frequency domain decomposition employs the Fast Fourier Decomposition algorithm, and the thermal cumulative oscillation coupling coefficient is obtained through the thermal cumulative oscillation coupling coefficient calculation formula, which is:

[0008] κ = η·δ·0.01;

[0009] In the formula, κ is the thermal accumulation oscillation coupling coefficient, η is the thermal accumulation attenuation coefficient, δ is the micro-oscillation attenuation coefficient, and 0.01 is the proportionality coefficient.

[0010] More preferably, the expression for the spatiotemporal heterogeneous attenuation correction formula for multimodal data is:

[0011] ;

[0012] In the formula, X * For the corrected single-mode basic parameter values, X is the original single-mode parameter value, α is the spatial attenuation factor, L is the distance between acquisition nodes, L0 is the distance between reference acquisition nodes, β is the time attenuation factor, τ is the data transmission delay, ω is the heterogeneous weight of data type, and λ is the heterogeneous coefficient of sampling frequency.

[0013] More preferably, the formula for calculating the thermal accumulation attenuation coefficient of the equipment is used to quantify the degree of attenuation of the adjustable capability due to the thermal accumulation of the power device based on the corrected thermal parameters of the equipment and the micro-oscillation characteristic parameters of the power grid. The formula for comprehensive prediction of the coupled adjustable capability is used to integrate the corrected basic operating parameters, the thermal accumulation attenuation coefficient and the thermal accumulation oscillation coupling coefficient to complete the prediction calculation of the active and reactive power adjustable capabilities.

[0014] More preferably, the corrected micro-oscillation angular frequency is obtained by converting the oscillation frequency in the corrected power grid oscillation parameters, the corrected micro-oscillation amplitude is consistent with the oscillation amplitude in the corrected power grid oscillation parameters, and the micro-oscillation attenuation coefficient is obtained by performing frequency domain decomposition on the corrected power grid oscillation parameters.

[0015] More preferably, the calculation process of the multimodal data spatiotemporal heterogeneous attenuation correction formula is as follows: the spatial attenuation factor, time attenuation factor, data type heterogeneous weight, sampling frequency heterogeneous coefficient and the original single-mode parameter value corresponding in the multimodal original dataset are substituted into the multimodal data spatiotemporal heterogeneous attenuation correction formula, and the calculations are performed on the electrical parameters, equipment thermal parameters and power grid oscillation parameters respectively to obtain the corrected electrical parameters, corrected equipment thermal parameters and corrected power grid oscillation parameters. The above corrected parameters are integrated to obtain the corrected basic operating parameter set.

[0016] Further preferably, the calculation process of the equipment thermal accumulation attenuation coefficient formula is as follows: extract the corrected device heat flux density and corrected micro-oscillation amplitude from the corrected basic operating parameter set; obtain the corrected device temperature rise by calculating the difference between the actual device temperature and the ambient temperature in the corrected device thermal parameters; substitute the corrected device heat flux density, corrected device temperature rise, corrected micro-oscillation amplitude, device heat dissipation area, thermal accumulation time, device material specific heat capacity, device heating core mass, power device thermal resistance coefficient, power grid micro-oscillation influence coefficient, and power grid micro-oscillation reference amplitude into the equipment thermal accumulation attenuation coefficient calculation formula, and sequentially complete the fractional operation, multiplication operation, and subtraction operation to obtain the thermal accumulation attenuation coefficient.

[0017] More preferably, the calculation process of the coupled adjustable capacity comprehensive prediction formula is as follows: First, the thermal cumulative oscillation coupling coefficient is obtained through the thermal cumulative oscillation coupling coefficient calculation formula. Then, the corrected active power, corrected reactive power, and corrected micro-oscillation angular frequency are extracted from the corrected basic operating parameter set. The corrected active power, thermal cumulative attenuation coefficient, thermal cumulative oscillation coupling coefficient, corrected micro-oscillation angular frequency, prediction time window, and distributed energy regulation response time are substituted into the active power adjustable capacity prediction formula to complete the calculation and obtain the active power adjustable capacity. At the same time, the corrected reactive power, thermal cumulative attenuation coefficient, thermal cumulative oscillation coupling coefficient, corrected micro-oscillation angular frequency, prediction time window, and distributed energy regulation response time are substituted into the reactive power adjustable capacity prediction formula to complete the calculation and obtain the reactive power adjustable capacity. The prediction time window is the prediction duration of the distributed energy adjustable capacity, and the distributed energy regulation response time is the power regulation response duration of the distributed energy after receiving the dispatch instruction.

[0018] Technical Effects: This invention precisely addresses the core issues in the background technology, namely, the failure to consider the multi-factor linkage effect and the incomplete prediction logic, through three innovative technical points: extracting and correcting spatiotemporal heterogeneous attenuation characteristic parameters, extracting power grid micro-oscillation characteristic parameters, and establishing a calculation model linking thermal accumulation and micro-oscillation. It improves the entire prediction logic, enables accurate prediction of the adjustable capability of distributed energy, adapts to complex operating scenarios, and ensures the reliability of power grid dispatch. Attached Figure Description

[0019] Figure 1 This is a flowchart of the distributed energy adjustability prediction method based on multimodal data fusion proposed in this application. Detailed Implementation

[0020] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention.

[0021] The existing technology has the following technical problems: it does not consider the linkage between the spatiotemporal heterogeneous decay of multimodal data and the thermal accumulation of equipment and the micro-oscillation of the power grid, and only conducts simple fusion prediction based on conventional parameters. The prediction logic is incomplete and cannot be adapted to complex operating scenarios.

[0022] Based on this, please refer to Figure 1 This embodiment provides a method for predicting the adjustable capacity of distributed energy resources based on multimodal data fusion, including the following steps:

[0023] S1: Collect multimodal raw datasets of distributed energy resources through multiple sensors, metering devices and power grid monitoring terminals. The multimodal raw datasets include electrical parameters, equipment thermal parameters, power grid oscillation parameters and spatiotemporal acquisition parameters.

[0024] S2: Extract the spatiotemporal heterogeneous decay feature parameters of the multimodal original dataset. The spatiotemporal heterogeneous decay feature parameters include spatial decay factor, temporal decay factor, data type heterogeneous weight, and sampling frequency heterogeneous coefficient. The multimodal original dataset is corrected using the spatiotemporal heterogeneous decay correction formula to obtain the corrected basic operating parameter set.

[0025] S3: Perform frequency domain decomposition on the grid oscillation parameters in the corrected basic operating parameter set to extract grid micro-oscillation characteristic parameters, including micro-oscillation amplitude, micro-oscillation angular frequency, and micro-oscillation attenuation coefficient;

[0026] S4: Substitute the corrected set of basic operating parameters and the power grid micro-oscillation characteristic parameters into the formula for calculating the equipment thermal cumulative attenuation coefficient to obtain the thermal cumulative attenuation coefficient;

[0027] S5: Calculate the thermal cumulative oscillation coupling coefficient based on the thermal cumulative attenuation coefficient and the micro-oscillation attenuation coefficient. Substitute the corrected basic operating parameter set, the thermal cumulative attenuation coefficient, and the thermal cumulative oscillation coupling coefficient into the coupled adjustable capability comprehensive prediction formula to obtain the active power adjustable capability and reactive power adjustable capability of distributed energy.

[0028] The technical solution described is as follows: the entire method constructs a complete prediction chain with multi-factor linkage quantification as its core. All technical features are essential components for solving core technical problems; the absence of any feature will lead to a break in the prediction logic, making it impossible to achieve progressive quantitative analysis of multi-dimensional influencing factors. First, four types of core parameters are acquired simultaneously through multi-source acquisition devices. This is the first time that spatiotemporal acquisition parameters have been incorporated into the basic data system for predicting the adjustable capability of distributed energy, filling the gap in existing technical data dimensions. Electrical parameters, equipment thermal parameters, and grid oscillation parameters respectively point to the core operation of distributed energy, equipment status, and grid environment. Spatiotemporal acquisition parameters point to the acquisition and transmission characteristics of the data itself. These four types of parameters form a comprehensive basic data system, providing data support for subsequent accurate calculations. Next, four core components of spatiotemporal heterogeneous attenuation characteristic parameters are extracted. These four components correspond to attenuation characteristics in four dimensions: space, time, data type, and sampling frequency. They are the core of achieving accurate multimodal data correction. Based on these characteristic parameters, a dedicated correction formula is used to perform data correction, eliminating parameter deviations caused by spatiotemporal transmission from the data source, providing accurate basic data for all subsequent calculation stages. Subsequently, frequency domain decomposition was performed on the corrected grid oscillation parameters to extract micro-oscillation-related characteristic parameters. This refined feature of grid oscillation was incorporated into the prediction system, filling the gap in existing technologies that did not consider the impact of grid micro-oscillations on equipment operation, and providing characteristic basis for the quantification of the linkage between heat accumulation and micro-oscillations. Finally, a three-step progressive calculation was used to achieve multi-factor linkage prediction. First, the heat accumulation attenuation coefficient was calculated based on the corrected parameters and micro-oscillation characteristic parameters to achieve preliminary linkage quantification between heat accumulation and micro-oscillations. Then, the coupling coefficient was calculated based on the heat accumulation attenuation coefficient and the micro-oscillation attenuation coefficient to accurately characterize the degree of synergistic influence between the two. Finally, the corrected basic parameters, heat accumulation attenuation coefficient, and coupling coefficient were substituted into the comprehensive prediction formula to achieve deep coupling quantification of multiple factors. The entire set of steps is progressive, with the result of the previous step serving as the sole input for the subsequent step, forming a seamless computational link. This progressive multi-factor linkage modeling method changes the single-dimensional and static prediction logic of existing technologies, realizing full-link optimization from data correction to parameter modeling to capacity prediction, allowing the prediction results to accurately match the actual adjustable capabilities of distributed energy in complex operating scenarios. The technical effects achieved by this solution are: to solve the problem of unconsidered multi-factor linkage effects, to improve the entire prediction logic, to adapt to complex operating scenarios, and to ensure the overall reliability of the prediction of the adjustable capability of distributed energy.

[0029] The existing technology has the following technical problems: the parameter types of the original multimodal dataset are not clearly defined, resulting in the lack of specificity of the collected data and failing to provide effective support for subsequent correction and prediction operations.

[0030] Based on this, the electrical parameters include active power, reactive power, and operating current; the equipment thermal parameters include device heat flux density, actual device temperature, and ambient temperature; the grid oscillation parameters include oscillation amplitude and oscillation frequency; and the spatiotemporal acquisition parameters include acquisition node distance, data transmission delay, and sampling frequency. The acquisition node distance is the physical distance between the distributed energy data acquisition node and the dispatch terminal; the data transmission delay is the time it takes for the multimodal raw data to be transmitted from the acquisition node to the dispatch terminal; and the sampling frequency is the acquisition frequency of the multimodal raw data. The technical solution of this plan is described as follows: based on the preceding multimodal raw dataset acquisition steps, it clearly defines the specific composition and definition of the four basic parameters, providing clear execution standards for the data acquisition process and ensuring that the acquired data accurately matches the parameter input requirements of subsequent correction formulas, attenuation coefficient calculation formulas, and comprehensive prediction formulas. Among the electrical parameters, active power, reactive power, and operating current are the core indicators for determining the adjustability of distributed energy resources. They are also the original values ​​for corrected active power and corrected reactive power in the subsequent comprehensive prediction formula, directly determining the core directionality of the prediction results. The thermal parameters selected—device heat flux density, actual device temperature, and ambient temperature—are key indicators characterizing the thermal accumulation state of power devices. The device heat flux density is the original value of the corrected device heat flux density in the calculation formula for the thermal accumulation attenuation coefficient. The difference between the actual device temperature and the ambient temperature is the basis for determining the normal device temperature rise in this calculation formula, directly affecting the accuracy of the thermal accumulation attenuation coefficient calculation. The selected oscillation amplitude and frequency parameters for power grid oscillation are the foundation for extracting the characteristics of power grid micro-oscillations. The oscillation amplitude is the original value for subsequent micro-oscillation amplitudes, and the oscillation frequency is the basis for converting the micro-oscillation angular frequency, which is the basis for realizing the quantitative linkage between power grid micro-oscillations and equipment thermal accumulation. The three specific indicators of the spatiotemporal acquisition parameters correspond to the values ​​of spatial attenuation, time attenuation, and sampling frequency heterogeneity in the correction formula, respectively. The acquisition node distance is the actual acquisition node distance in the correction formula, the data transmission delay is the transmission delay in the formula, and the sampling frequency is the basis for calculating the sampling frequency heterogeneity coefficient. These three types of indicators jointly determine the calculation result of the correction formula. At the same time, the specific definitions of the spatiotemporal acquisition parameters are clearly defined to avoid deviations in the parameter values ​​of the correction formula due to ambiguity in the judgment range of acquisition node distance and transmission delay, ensuring the accuracy of the calculation in the correction process. The technical effect achieved by this solution is: clarifying the specific composition and value basis of multimodal data, providing targeted data support for the accurate calculation of various subsequent formulas, and reducing calculation deviations caused by parameter ambiguity.

[0031] The existing technology has the following technical problems: the composition and source of the basic operating parameter set after correction are not clear, which leads to poor connection between the correction stage and the subsequent calculation stage and cannot guarantee data consistency.

[0032] Based on this, the corrected basic operating parameter set includes corrected electrical parameters, corrected equipment thermal parameters, and corrected grid oscillation parameters. The corrected electrical parameters are obtained by calculating the electrical parameters using the multimodal data spatiotemporal heterogeneous attenuation correction formula. The corrected equipment thermal parameters are obtained by calculating the equipment thermal parameters using the multimodal data spatiotemporal heterogeneous attenuation correction formula. The corrected grid oscillation parameters are obtained by calculating the grid oscillation parameters using the multimodal data spatiotemporal heterogeneous attenuation correction formula. The technical solution of this approach connects to the preceding multimodal data correction steps, clearly defining the specific composition and source of the corrected basic operating parameter set. This achieves seamless integration between the correction stage and subsequent stages such as thermal accumulation attenuation coefficient calculation, micro-oscillation feature extraction, and comprehensive prediction, ensuring the consistency and standardization of the entire computational data chain. The three categories of the corrected basic operating parameter set correspond one-to-one with the three core parameters of the original multimodal dataset, ensuring that the corrected parameters accurately match the computational needs of subsequent stages. The corrected electrical parameters directly provide the corrected active power and reactive power values ​​for the comprehensive prediction formula. The corrected equipment thermal parameters provide the corrected device heat flux density and temperature rise values ​​for the calculation formula of the heat accumulation attenuation coefficient. The corrected grid oscillation parameters provide the basis for micro-oscillation feature extraction, and thus provide the values ​​of micro-oscillation-related parameters for the calculation formula of the heat accumulation attenuation coefficient and the comprehensive prediction formula. Furthermore, it is clarified that all three types of corrected parameters are calculated using the same spatiotemporal heterogeneous attenuation correction formula for multimodal data, ensuring a unified correction standard for all corrected parameters and avoiding data deviations caused by different correction methods for different parameters. This ensures that all subsequent calculations based on the corrected parameters maintain consistency in the data benchmark. This one-to-one parameter correction relationship and unified correction formula application principle make the correction stage a unified data source for all subsequent calculation stages. The calculation result of the preceding correction formula directly determines the input accuracy of all subsequent formulas, forming a data calibration system centered on the correction formula. This fills the gap in existing technologies where there is no unified correction for multimodal data and the data benchmark is chaotic. Furthermore, the clear definition of parameter sources makes the calculation logic of the entire method clearer. The parameters for each type of subsequent calculation can be traced back to the original collected data and the calculation process of the correction formula, facilitating subsequent result verification and parameter adjustment. This makes the entire method operable and optimizable in practical applications. The technical effects achieved by this solution are: clarifying the composition and source of the corrected data, ensuring the smooth connection of data in each calculation stage, and improving the accuracy and standardization of the entire calculation chain.

[0033] The existing technology has the following technical problems: the frequency domain decomposition algorithm is not clear and the logic for calculating the coupling coefficient is ambiguous, which leads to inaccurate extraction of power grid micro-oscillation features and lack of basis for calculating the coupling coefficient, thus affecting the reliability of the prediction results.

[0034] Based on this, the frequency domain decomposition employs the Fast Fourier Decomposition algorithm, and the thermal cumulative oscillation coupling coefficient is obtained through the thermal cumulative oscillation coupling coefficient calculation formula, which is:

[0035] ;

[0036] In the formula The thermal accumulation oscillation coupling coefficient is... The thermal accumulation decay coefficient, This is the micro-oscillation attenuation coefficient. This refers to the proportionality coefficient. The technical solution outlines the detailed steps for extracting power grid micro-oscillation features and calculating the thermal accumulation oscillation coupling coefficient, filling operational and logical gaps in existing technologies at this stage. This provides clear implementation standards and calculation basis for the deep coupling quantification of thermal accumulation and power grid micro-oscillations. The Fast Fourier Decomposition (FFT) algorithm is selected for power grid micro-oscillation feature extraction. This algorithm is characterized by high computational efficiency and accurate feature decomposition, enabling rapid extraction of three core feature parameters—micro-oscillation amplitude, micro-oscillation angular frequency, and micro-oscillation attenuation coefficient—from the corrected power grid oscillation parameters. Furthermore, the accuracy of the decomposition results matches the computational requirements of subsequent formulas, avoiding linkage quantization deviations caused by inaccurate feature extraction. This is crucial for achieving refined feature analysis of power grid micro-oscillations. Simultaneously, a specific formula for calculating the thermal accumulation oscillation coupling coefficient is defined. This formula is the core for achieving deep coupling quantification of thermal accumulation and grid micro-oscillation. In the formula, the thermal accumulation attenuation coefficient is the result of the calculation formula for the equipment's thermal accumulation attenuation coefficient, and the micro-oscillation attenuation coefficient is the result extracted by the Fast Fourier Decomposition algorithm. Multiplying these two core parameters, which respectively characterize the thermal state of the equipment and the micro-oscillation state of the grid, can accurately characterize the degree of synergistic influence between the two. The proportionality coefficient... This is used to normalize the product result, controlling the value range of the coupling coefficient within a reasonable range. This avoids the coupling coefficient from being too large or too small due to differences in the values ​​of the two coefficients, which could affect the calculation result of the comprehensive prediction formula and ensure that the coupling coefficient accurately reflects the actual coupling degree between the two factors. The design of this calculation formula follows the theoretical basis of normalization modeling, quantifying the attenuation parameters of two different dimensions in a linked manner. This changes the logic of existing technologies that calculate equipment heat accumulation and grid oscillation separately, achieving accurate quantification of the coupling degree between the two. The technical effects achieved by this solution are: clarifying the frequency domain decomposition algorithm and the coupling coefficient calculation logic, improving the accuracy of micro-oscillation feature extraction and the reliability of coupling coefficient calculation, and providing a basis for multi-factor coupling prediction.

[0037] The existing technology has the following technical problems: the spatiotemporal heterogeneous attenuation correction of multimodal data lacks a dedicated quantification formula, which cannot accurately eliminate the deviation of the original data caused by spatiotemporal transmission, thus affecting the accuracy of subsequent calculations.

[0038] Based on this, the expression for the spatiotemporal heterogeneous attenuation correction formula for multimodal data is:

[0039] ;

[0040] In the formula To correct the basic parameter values ​​for a single mode, These are the original single-mode parameter values. The spatial decay factor, To collect node distance, The distance between the baseline data acquisition nodes The time decay factor, For data transmission latency, For heterogeneous weights of data types, The sampling frequency heterogeneity coefficient is described as follows: The technical solution is a quantization correction formula specifically designed for the spatiotemporal heterogeneous attenuation characteristics of multimodal data. The entire formula is based on the signal exponential attenuation model in the communication field and the heterogeneous weighting theory of multimodal data fusion. It is also specifically optimized to suit the actual characteristics of distributed energy data acquisition and transmission. This design conforms to the basic laws of signal transmission and accurately adapts to the heterogeneous attenuation correction requirements of multimodal data. The selection of all parameters and the design of the computational logic in the formula revolve around attenuation correction in the spatiotemporal and heterogeneous dimensions, achieving precise nonlinear correction of the original multimodal data. This replaces the simple linear weighted fusion method in existing technologies, eliminating parameter deviations caused by spatiotemporal transmission and heterogeneous characteristics from the data source. In the formula… and For single-mode parameter values ​​of the same type, it can be adapted to the correction of any type of parameter, such as electrical parameters, equipment thermal parameters, and power grid oscillation parameters, making the formula universal and enabling unified correction of multi-mode data, ensuring the consistency of the corrected data; The spatial attenuation factor, a dimensionless parameter, is determined by the actual scenario of distributed energy data acquisition and is used to characterize the degree of impact of spatial distance on data attenuation under different acquisition scenarios. and Both are length units. The ratio of the two achieves dimensionless processing of spatial attenuation-related parameters, avoiding the influence of dimensions on the exponential function calculation. At the same time, it allows the spatial attenuation calculation results to fit the ratio of the actual acquisition node distance to the reference distance, accurately representing the degree of attenuation caused by spatial distance. As a time decay factor, its dimension is the reciprocal of a second, and it is related to the dimension of time. Multiplication achieves dimensionless representation of time decay-related parameters. The value of is determined by the characteristics of the communication link in data transmission, accurately characterizing the impact of latency on data attenuation under different transmission links. The exponential function's calculation form closely reflects the actual nonlinear attenuation of signals in spatiotemporal transmission. Compared to the linear weighting method, it more accurately reflects the actual attenuation degree under different spatiotemporal distances and delays. The result of the exponential function is a value between 0 and 1, and multiplying it by the original parameters can achieve attenuation correction for the original parameters. The formula contains... and All are dimensionless parameters, representing the heterogeneous attenuation effects caused by data type and sampling frequency, respectively. The values ​​are determined based on the attenuation characteristics of different data types, including sensing, communication, and metering. The heterogeneous attenuation correction weight is determined by the ratio of the actual sampling frequency to the reference sampling frequency. The product of these two frequencies, plus 1 and taking the reciprocal, forms the weight. This weight, also a value between 0 and 1, enables differentiated heterogeneous attenuation correction based on data type and sampling frequency, filling the gap in existing technologies that do not consider heterogeneous attenuation in multimodal data. The entire formula combines nonlinear correction of spatiotemporal attenuation with differentiated correction of heterogeneous attenuation. Two dimensionless transformations ensure the homogeneity of the formula's dimensions; all composite terms are dimensionless. and With completely consistent dimensions, the formula's calculation results not only conform to physical laws but also accurately eliminate various attenuation deviations in multimodal data acquisition and transmission. Each parameter in the formula has a clear physical meaning and value basis, and each calculation step has corresponding theoretical support. This achieves accurate and unified correction of multimodal data, providing precise foundational data for all subsequent calculation stages. This is the core innovation of this formula compared to the simple weighted fusion method of existing technologies: by combining nonlinear correction and heterogeneous differential correction, it solves the industry pain point of inaccurate correction of spatiotemporal heterogeneous attenuation in multimodal data. The technical effects achieved by this solution are: providing a dedicated multimodal data correction formula, accurately eliminating parameter deviations caused by spatiotemporal heterogeneous attenuation, ensuring the accuracy of the original data, and laying a precise data foundation for subsequent end-to-end calculations.

[0041] The existing technology has the following technical problems: the comprehensive prediction of equipment thermal accumulation decay and adjustability lacks a clear definition of the calculation logic, resulting in inaccurate quantification of the impact of thermal accumulation and no clear basis for comprehensive prediction.

[0042] Based on this, the formula for calculating the equipment thermal accumulation attenuation coefficient is used to quantify the degree of attenuation of the adjustability due to thermal accumulation of power devices, based on the corrected equipment thermal parameters and the micro-oscillation characteristic parameters of the power grid. The expression for the formula for calculating the equipment thermal accumulation attenuation coefficient is as follows:

[0043] ;

[0044] In the formula The cumulative thermal decay coefficient of the equipment. The thermal resistance coefficient of the power device. To correct the heat flux density of the device, For the heat dissipation area of ​​the device, For heat accumulation time, For the specific heat capacity of the device material, For the core quality of the device's heat dissipation, To correct for the temperature rise of the device, The influence coefficient of power grid micro-oscillation. To correct the amplitude of micro-oscillations in the power grid, As the reference amplitude for power grid micro-oscillations, the coupled adjustable capability comprehensive prediction formula is used to integrate the corrected basic operating parameters, thermal accumulation attenuation coefficient, and thermal accumulation oscillation coupling coefficient to complete the prediction calculation of active and reactive power adjustable capabilities. The technical solution of this scheme clarifies the core uses and operational logic of two types of core calculation formulas. It also discloses for the first time a dedicated quantitative formula for the equipment thermal accumulation attenuation coefficient. This formula is based on Fourier's law of heat conduction in thermodynamics and the thermal accumulation model of power electronic devices. It combines the actual impact of power grid micro-oscillations on device current fluctuations for targeted modeling, overcoming the limitations of existing technologies that only calculate equipment thermal accumulation based on pure thermodynamics. This achieves preliminary linkage quantification between thermal accumulation and power grid micro-oscillations. All core input parameters in the formula... , , All results are calculated using the spatiotemporal heterogeneous attenuation correction formula for preceding multimodal data. This achieves a progressive linkage with the correction process, ensuring that the calculation of the thermal accumulation attenuation coefficient is based on precisely corrected parameters, thus fundamentally improving quantization accuracy. In the formula... This is a dimensionless thermal accumulation attenuation coefficient for devices, ranging from 0 to 1. Its physical meaning is the proportion of adjustable capability remaining after thermal accumulation of a power device. It is a dimensionless thermal resistivity of power devices, determined by the material and process characteristics of the device, and is used to characterize the degree to which different device materials impede heat accumulation. The corrected heat flux density of the device is expressed in watts per square meter. The heat dissipation area of ​​the device is measured in square meters. The thermal accumulation time is measured in seconds. The product of these three values ​​is the total thermal energy of the device during the thermal accumulation time, measured in joules, which conforms to the energy calculation logic of Fourier's law of thermal conduction. The specific heat capacity of the device material is expressed in joules per kilogram in Kelvin. The mass of the heat-generating core of the device is measured in kilograms. The corrected temperature rise of the device is expressed in Kelvin. The product of the three is the heat required for the device to reach that temperature rise, which is also expressed in Joules. This is consistent with the dimension of the thermal energy of molecules. The ratio of the two is the dimensionless degree of heat accumulation, which accurately characterizes the relative relationship between the actual heat accumulation of the device and its own temperature rise carrying capacity. This is a dimensionless power grid micro-oscillation influence coefficient, determined by the actual operating characteristics of the power grid, used to characterize the degree to which micro-oscillations enhance heat accumulation. and Both are oscillation amplitudes with dimensions in volts. The ratio of the two achieves dimensionlessness, allowing the influence of micro-oscillations to be quantified and incorporated into the heat accumulation calculation. The term represents the heat accumulation enhancement factor in the presence of micro-oscillations. The larger the amplitude of the micro-oscillation, the larger the value of this factor, and the higher the degree of heat accumulation. This aligns with the actual physical law that micro-oscillations in the power grid exacerbate current fluctuations in devices, thereby accelerating heat accumulation. The overall calculation logic of the formula follows the steps of quantifying the degree of heat accumulation, correcting for micro-oscillation enhancement, and deriving the attenuation coefficient. First, the basic degree of heat accumulation is obtained through fractional operations. Then, the micro-oscillation enhancement factor is introduced for linkage correction. Finally, the final attenuation coefficient is obtained through multiplication and subtraction with the thermal resistance coefficient of the power device. All composite terms are dimensionless, ensuring... The dimensionless nature of the formula conforms to the definition specifications of physical quantities. The core innovation of this formula lies in incorporating grid micro-oscillations as an enhancing factor of thermal accumulation into the quantification model, while simultaneously achieving progressive parameter linkage with the preceding correction formula. This addresses the industry pain point that existing technologies cannot quantify the linkage effect between grid micro-oscillations and equipment thermal accumulation. The coupled adjustable capability comprehensive prediction formula serves as the core carrier for deep multi-factor coupling, receiving the calculation results of the thermal accumulation attenuation coefficient, and achieving a logical connection from single-parameter quantification to multi-factor coupled prediction. The two types of formulas form a progressive operational system, together with the preceding correction formula, constituting a seamless end-to-end modeling system. The technical effects achieved by this solution are: clarifying the purpose and operational logic of the core calculation formula, providing a dedicated thermal accumulation attenuation quantification formula, improving the accuracy of thermal accumulation impact quantification and the standardization of comprehensive prediction, and perfecting the end-to-end logic of multi-factor linkage quantification.

[0045] The existing technology has the following technical problems: the source and conversion logic of the corrected micro-oscillation characteristic parameters are not clear, which leads to poor connection between the micro-oscillation characteristic parameters and the corrected power grid oscillation parameters, affecting the accuracy of linkage calculation.

[0046] Based on this, the corrected micro-oscillation angular frequency is calculated from the oscillation frequency in the corrected power grid oscillation parameters, the corrected micro-oscillation amplitude is consistent with the oscillation amplitude in the corrected power grid oscillation parameters, and the micro-oscillation attenuation coefficient is obtained by frequency domain decomposition of the corrected power grid oscillation parameters. The technical solution is described as follows: this content connects with the preceding power grid micro-oscillation feature extraction steps, clearly defining the sources and conversion logic of the three types of micro-oscillation feature parameters, achieving seamless connection between the corrected power grid oscillation parameters and the micro-oscillation feature parameters, ensuring the accuracy of the parameters for the quantification of power grid micro-oscillation and equipment thermal accumulation. These three types of micro-oscillation feature parameters are also key input parameters for the calculation formula of the equipment thermal accumulation attenuation coefficient and the comprehensive prediction formula for coupled adjustable capability; their accuracy directly determines the effect of multi-factor linkage quantification. The corrected micro-oscillation amplitude directly adopts the oscillation amplitude from the corrected grid oscillation parameters, ensuring that this parameter is consistent with the corrected grid oscillation parameters on the data basis, avoiding parameter deviations caused by secondary conversion. This parameter is the direct basis for determining the normal micro-oscillation amplitude in the formula for calculating the equipment thermal accumulation attenuation coefficient, and directly affects the quantitative accuracy of the grid micro-oscillation's enhancement effect on equipment thermal accumulation. The corrected micro-oscillation angular frequency is obtained by converting the oscillation frequency from the corrected grid oscillation parameters. The conversion logic conforms to the basic conversion rules between angular frequency and frequency in the power system, ensuring that the value of this parameter conforms to physical laws. This parameter is the core parameter characterizing the periodicity of grid micro-oscillation in the coupled adjustable capability comprehensive prediction formula, and directly determines the calculation result of the oscillation correction term in the formula. The micro-oscillation attenuation coefficient is obtained by performing frequency domain decomposition on the corrected grid oscillation parameters, echoing the previous frequency domain decomposition step, ensuring that the extraction process of this parameter is consistent with the corrected grid oscillation parameters. This parameter is one of the core inputs in the formula for calculating the thermal accumulation oscillation coupling coefficient, and directly determines the quantitative accuracy of the coupling degree between thermal accumulation and grid micro-oscillation. Meanwhile, all three types of micro-oscillation characteristic parameters are derived from the corrected grid oscillation parameters, forming a data connection with the preceding multi-modal data correction stage. This ensures that the data source of the micro-oscillation characteristic parameters remains consistent with other parameters in subsequent calculations, avoiding calculation deviations caused by different data sources. This allows the characteristic parameters of grid micro-oscillations to accurately match the relevant parameters of equipment thermal accumulation, achieving precise linkage quantification between the two. The technical effects achieved by this solution are: clarifying the source and conversion logic of micro-oscillation characteristic parameters, ensuring smooth parameter connection, and improving the accuracy of the linkage calculation between grid micro-oscillations and equipment thermal accumulation.

[0047] The existing technology has the following technical problems: the calculation process of the spatiotemporal heterogeneous attenuation correction formula for multimodal data is not clear, resulting in no standardized procedure for correction operation, which is prone to calculation deviation and affects the quality of the corrected data.

[0048] Based on this, the calculation process of the spatiotemporal heterogeneous attenuation correction formula for multimodal data is as follows: Substitute the spatial attenuation factor, temporal attenuation factor, data type heterogeneous weight, sampling frequency heterogeneous coefficient, and the corresponding original single-mode parameter values ​​in the original multimodal dataset into the spatiotemporal heterogeneous attenuation correction formula. Perform calculations on electrical parameters, equipment thermal parameters, and power grid oscillation parameters respectively to obtain the corrected electrical parameters, corrected equipment thermal parameters, and corrected power grid oscillation parameters. Integrate these corrected parameters to obtain the corrected basic operating parameter set. The technical solution of this method clarifies the standardized calculation process of the spatiotemporal heterogeneous attenuation correction formula for multimodal data, providing a clear basis for the correction operation, avoiding calculation deviations caused by non-standard operation procedures, and ensuring the quality and consistency of the corrected data. Furthermore, the design of this calculation process is consistent with the physical meaning and calculation logic of the formula, maximizing the correction effect of the formula. The first step in the calculation process is parameter matching and substitution. The four core components of the spatiotemporal heterogeneous attenuation characteristic parameters are matched one-to-one with the original single-mode parameter values ​​and substituted into the formula. This ensures that each parameter in the formula has a precise value, avoiding calculation deviations caused by incorrect parameter substitution. This step is fundamental to ensuring the accuracy of the formula calculation. The second step is categorized parameter calculation. Formula calculations are performed separately for electrical parameters, equipment thermal parameters, and power grid oscillation parameters. Since the physical meanings and dimensions of these three types of parameters differ, categorized calculation avoids logical confusion and result deviations caused by mixing parameters of different dimensions and types. Simultaneously, it ensures that the correction result for each type of parameter accurately matches its own characteristics. The physical characteristics of the device, such as the correction results of electrical parameters, can accurately reflect its actual operating value after spatiotemporal heterogeneous attenuation. The correction results of equipment thermal parameters can accurately reflect the actual thermal state of the device. Categorical calculation is also the key to achieving unified correction of multimodal data while maintaining the independence of various parameters. The third step is the integration of corrected parameters. The corrected electrical parameters, corrected equipment thermal parameters, and corrected grid oscillation parameters obtained from categorical calculation are integrated into a set of corrected basic operating parameters, forming a unified corrected data system. This provides unified parameter input for subsequent micro-oscillation feature extraction, thermal accumulation attenuation coefficient calculation, and comprehensive prediction, achieving seamless connection between the correction and subsequent calculation stages. This standardized calculation process makes the application of the correction formula operable and replicable. Consistent and accurate correction results can be obtained under different distributed energy acquisition scenarios and different operators. This solves the problems of lack of standardization and poor consistency of results in existing technology correction operations. At the same time, the calculation process matches the progressive calculation logic of the formula, allowing the spatiotemporal heterogeneous attenuation correction effect of the formula to be fully implemented, providing a guarantee for accurate calculation of the entire chain. The technical effect achieved by this solution is to standardize the calculation process of the correction formula, reduce calculation deviations caused by non-standard operation, and improve the quality and consistency of the corrected data.

[0049] The existing technology has the following technical problems: the calculation process of the equipment thermal cumulative decay coefficient is not clear, resulting in no standardized process for calculating the thermal cumulative decay coefficient, which is prone to calculation errors and affects the subsequent coupled calculation and prediction results.

[0050] Based on this, the calculation process of the equipment thermal accumulation attenuation coefficient formula is as follows: Extract the corrected device heat flux density and corrected micro-oscillation amplitude from the corrected basic operating parameter set; calculate the temperature rise of the corrected device by subtracting the actual device temperature from the ambient temperature in the corrected device thermal parameters; substitute the corrected device heat flux density, corrected device temperature rise, corrected micro-oscillation amplitude, device heat dissipation area, thermal accumulation time, device material specific heat capacity, device heating core mass, power device thermal resistance coefficient, power grid micro-oscillation influence coefficient, and power grid micro-oscillation reference amplitude into the equipment thermal accumulation attenuation coefficient calculation formula, and sequentially perform fractional operations, multiplication operations, and subtraction operations to obtain the thermal accumulation attenuation coefficient. The technical solution of this method clarifies the standardized calculation process of the equipment thermal accumulation attenuation coefficient formula. The design of this process strictly follows the mathematical operation logic of the formula and the basic laws of thermodynamics, ensuring the accuracy and reliability of the calculated thermal accumulation attenuation coefficient, while also making the calculation process operable and avoiding calculation errors caused by incorrect operation order or parameter extraction deviations.The first step in the calculation process is the extraction and calculation of core parameters. The corrected device heat flux density and corrected micro-oscillation amplitude are accurately extracted from the set of corrected basic operating parameters. These two parameters are the core inputs for realizing the quantification of the linkage between heat accumulation and grid micro-oscillation. Simultaneously, the corrected device temperature rise is obtained by calculating the difference between the actual device temperature and the ambient temperature in the corrected device thermal parameters. This parameter is a core indicator characterizing the device's heat accumulation state. The accuracy of parameter extraction and calculation directly determines the basis for subsequent formula calculations. All extracted parameters are the results of calculations in the preceding correction formulas, ensuring parameter accuracy. The second step is full parameter substitution. The three types of parameters obtained from extraction and calculation, along with inherent device parameters such as device heat dissipation area and heat accumulation time, and quantification coefficients such as power device thermal resistance coefficient and grid micro-oscillation influence coefficient, are all substituted into the calculation formula. This ensures that every parameter in the formula has a precise value. The inherent device parameters are determined by the production standards of distributed energy power devices, and the quantification coefficients are determined by the actual operating scenario and grid characteristics. All parameter values ​​are... There is clear evidence; the third step is a step-by-step calculation, which is carried out in the order of fractional calculation, product calculation, and subtraction calculation. This calculation order strictly follows the basic laws of mathematical operations and the modeling logic of formulas. First, the calculation of the corrected device heat flux density, device heat dissipation area, heat accumulation time and device material specific heat capacity, device heating core mass, and corrected device temperature rise is completed in the fractional part. The result of this part of the calculation is a dimensionless value characterizing the basic heat accumulation degree of the device. Then, this result is multiplied with the thermal resistance coefficient of the power device, the influence coefficient of the power grid micro-oscillation, and the ratio of the corrected micro-oscillation amplitude to the reference amplitude of the power grid micro-oscillation. This realizes the linkage quantification of the basic heat accumulation degree and the power grid micro-oscillation enhancement effect, so that the calculation result fits the actual physical law of the power grid micro-oscillation aggravating heat accumulation. Finally, the result of the product operation is subtracted from 1 to obtain the final device heat accumulation attenuation coefficient. This step-by-step calculation method can avoid the result deviation caused by the incorrect calculation order, and at the same time make the physical meaning of each step of the calculation clearer, which facilitates the verification of calculation results and parameter adjustment. Each step of this calculation process aligns with the design logic of the device's thermal accumulation attenuation coefficient calculation formula. This allows the formula to fully leverage the quantification effect of the linkage between thermal accumulation and grid micro-oscillations. The calculated thermal accumulation attenuation coefficient accurately characterizes the degree of attenuation of the power device's thermal accumulation on the adjustable capability of distributed energy resources under grid micro-oscillation environments. This provides precise parameter support for subsequent coupling coefficient calculations and comprehensive prediction formula calculations. Furthermore, this process makes the calculation of the thermal accumulation attenuation coefficient operable, addressing the industry pain point of the lack of a standardized process for thermal accumulation quantification in existing technologies. The technical effects achieved by this solution are: standardizing the calculation process of the thermal accumulation attenuation coefficient formula, reducing calculation errors, and improving the accuracy and reliability of the thermal accumulation attenuation coefficient calculation.

[0051] The existing technology has the following technical problems: the calculation process of the coupled adjustable capacity comprehensive prediction formula is not clear, resulting in no standardized process for predicting active and reactive power adjustable capacity, which is prone to calculation deviation and affects the reliability of prediction results.

[0052] Based on this, the calculation process of the coupled adjustable capacity comprehensive prediction formula is as follows: First, the thermal accumulation oscillation coupling coefficient is obtained through the thermal accumulation oscillation coupling coefficient calculation formula. Then, the corrected active power, corrected reactive power, and corrected micro-oscillation angular frequency are extracted from the corrected basic operating parameter set. The corrected active power, thermal accumulation attenuation coefficient, thermal accumulation oscillation coupling coefficient, corrected micro-oscillation angular frequency, prediction time window, and distributed energy regulation response time are substituted into the active power adjustable capacity prediction formula to complete the calculation and obtain the active power adjustable capacity. At the same time, the corrected reactive power, thermal accumulation attenuation coefficient, thermal accumulation oscillation coupling coefficient, corrected micro-oscillation angular frequency, prediction time window, and distributed energy regulation response time are substituted into the reactive power adjustable capacity prediction formula to complete the calculation and obtain the reactive power adjustable capacity. The expression of the active power adjustable capacity prediction formula is as follows:

[0053] ;

[0054] The expression for the reactive power adjustability prediction formula is as follows:

[0055] ;

[0056] In the formula For distributed energy resources with adjustable active power capability, For the corrected active power, The cumulative thermal decay coefficient of the equipment. The thermal accumulation oscillation coupling coefficient is... To correct the angular frequency of the micro-oscillation in the power grid, To predict the time window, For distributed energy regulation response time, For distributed energy resources, reactive power adjustment capability, This is for the corrected reactive power. The technical solution outlines a standardized calculation process for the coupled adjustable capacity comprehensive prediction formula, and fully discloses the dedicated prediction formulas for active and reactive power adjustable capacity. These two types of formulas are based on the fundamental calculation model of distributed energy adjustable capacity, the coupled modeling theory of control engineering, and the power oscillation law of power systems. They achieve deep coupling quantification of multiple factors, including thermal accumulation attenuation, thermal accumulation-grid micro-oscillation coupling, and the periodic influence of grid micro-oscillations. All core input parameters in the formulas... , Derived from the preceding correction formula, From the formula for calculating the cumulative thermal decay coefficient of equipment, From the formula for calculating the thermal accumulation oscillation coupling coefficient, Derived from the preceding frequency domain decomposition step, it achieves progressive linkage with all preceding stages, constructing a full-link modeling system from data correction to multi-factor coupled prediction. In the formula... The dimension is watt. The dimensions are finite, which aligns with the physical quantities of active and reactive power, respectively. and The selection of trigonometric functions is based on the different impact characteristics of power grid micro-oscillations on active and reactive power. The impact of power grid micro-oscillations on active power exhibits a sinusoidal periodic change, while the impact on reactive power exhibits a cosine periodic change. This design makes the prediction formula more in line with the actual operating law of the power system and is one of the core creative designs of the formula. The actual frequency of the micro-oscillation in the power grid, with dimensions in Hertz. To make the ratio of the prediction time window to the adjustment response time dimensionless, the product of the two makes the independent variable of the trigonometric function dimensionless in radians, which conforms to the norms of mathematical operations. This is the thermal accumulation oscillation coupling coefficient, used to characterize the degree of deep synergistic influence between thermal accumulation and micro-oscillations. Multiplying it by a trigonometric function term forms the coupled oscillation correction term. The form of the correction term ensures that the adjustable capability is always positive, consistent with physical meaning. The calculation process of this formula strictly follows the logic of pre-calculating the coupling coefficient, extracting core parameters, and simultaneously calculating sub-formulas. First, the coupling coefficient is calculated using the thermal accumulation oscillation coupling coefficient formula. This provides core quantitative basis for coupled prediction, and then extracts data from the corrected set of basic operating parameters. , , By selecting core parameters and then substituting them into the active and reactive power prediction formulas for synchronous calculations, the prediction efficiency is improved, and the prediction results for active and reactive power are tailored to the different impact characteristics of micro-oscillations in the power grid. This calculation process is highly consistent with the modeling logic of the coupled adjustable capacity comprehensive prediction formula, allowing the prediction effect of deep coupling of multiple factors to be fully realized. The calculated active and reactive power adjustable capabilities can accurately reflect the actual state of distributed energy resources under complex operating scenarios, solving the industry pain point that existing technologies can only make single-dimensional predictions based on static parameters. The technical effects achieved by this solution are: standardizing the calculation process of the comprehensive prediction formula, reducing calculation errors, and improving the accuracy and reliability of the prediction of the active and reactive power adjustable capabilities of distributed energy resources.

[0057] The above are merely preferred embodiments of the present invention and are not intended to limit the present invention in any other way. Any person skilled in the art may make changes or modifications to the above-disclosed technical content to create equivalent embodiments that can be applied to other fields. However, any simple modifications, equivalent changes, and modifications made to the above embodiments based on the technical essence of the present invention without departing from the scope of the present invention shall still fall within the protection scope of the present invention.

Claims

1. A method for predicting the adjustable capacity of distributed energy resources based on multimodal data fusion, characterized in that, Includes the following steps: Multimodal raw datasets of distributed energy resources are collected through multiple sensors, metering devices, and power grid monitoring terminals. These datasets include electrical parameters, equipment thermal parameters, power grid oscillation parameters, and spatiotemporal acquisition parameters. Spatiotemporal heterogeneous attenuation characteristic parameters are extracted from the datasets, including spatial attenuation factor, temporal attenuation factor, data type heterogeneity weight, and sampling frequency heterogeneity coefficient. The multimodal raw datasets are then corrected using a multimodal data spatiotemporal heterogeneous attenuation correction formula to obtain a corrected set of basic operating parameters. The expression for the multimodal data spatiotemporal heterogeneous attenuation correction formula is as follows: In the formula, To correct the basic parameter values ​​for a single mode, These are the original single-mode parameter values. The spatial decay factor, To collect node distance, The distance between the baseline data acquisition nodes The time decay factor, For data transmission latency, For heterogeneous weights of data types, The sampling frequency heterogeneity coefficient is used; the power grid oscillation parameters in the corrected basic operating parameter set are decomposed in the frequency domain to extract the power grid micro-oscillation characteristic parameters, which include micro-oscillation amplitude, micro-oscillation angular frequency, and micro-oscillation attenuation coefficient; Substituting the corrected basic operating parameter set and the power grid micro-oscillation characteristic parameters into the equipment thermal cumulative attenuation coefficient calculation formula, the thermal cumulative attenuation coefficient is obtained; based on the thermal cumulative attenuation coefficient and the micro-oscillation attenuation coefficient, the thermal cumulative oscillation coupling coefficient is calculated; substituting the corrected basic operating parameter set, the thermal cumulative attenuation coefficient, and the thermal cumulative oscillation coupling coefficient into the coupled adjustable capability comprehensive prediction formula, the active power adjustable capability and reactive power adjustable capability of distributed energy are obtained.

2. The method for predicting the adjustable capacity of distributed energy resources based on multimodal data fusion according to claim 1, characterized in that, The electrical parameters include active power, reactive power, and operating current; the equipment thermal parameters include device heat flux density, actual device temperature, and ambient temperature; the grid oscillation parameters include oscillation amplitude and oscillation frequency; and the spatiotemporal acquisition parameters include acquisition node distance, data transmission delay, and sampling frequency. The acquisition node distance is the physical distance between the distributed energy data acquisition node and the dispatch terminal; the data transmission delay is the time it takes for multimodal raw data to be transmitted from the acquisition node to the dispatch terminal; and the sampling frequency is the acquisition frequency of the multimodal raw data.

3. The method for predicting the adjustable capacity of distributed energy resources based on multimodal data fusion according to claim 1, characterized in that, The corrected basic operating parameter set includes corrected electrical parameters, corrected equipment thermal parameters, and corrected power grid oscillation parameters. The corrected electrical parameters are obtained by calculating the electrical parameters using the multimodal data spatiotemporal heterogeneous attenuation correction formula. The corrected equipment thermal parameters are obtained by calculating the equipment thermal parameters using the multimodal data spatiotemporal heterogeneous attenuation correction formula. The corrected power grid oscillation parameters are obtained by calculating the power grid oscillation parameters using the multimodal data spatiotemporal heterogeneous attenuation correction formula.

4. The method for predicting the adjustable capacity of distributed energy resources based on multimodal data fusion according to claim 1, characterized in that, The frequency domain decomposition employs the Fast Fourier Decomposition algorithm, and the thermal cumulative oscillation coupling coefficient is obtained through the thermal cumulative oscillation coupling coefficient calculation formula, which is: ； In the formula The thermal accumulation oscillation coupling coefficient is... The thermal accumulation decay coefficient, is the micro-oscillation attenuation coefficient, and 0.01 is the proportionality coefficient.

5. The method for predicting the adjustable capacity of distributed energy resources based on multimodal data fusion according to claim 1, characterized in that, The formula for calculating the thermal accumulation attenuation coefficient of the equipment is used to quantify the degree of attenuation of the adjustable capability due to thermal accumulation of power devices based on the corrected thermal parameters of the equipment and the micro-oscillation characteristic parameters of the power grid. The formula for comprehensive prediction of the coupled adjustable capability is used to integrate the corrected basic operating parameters, the thermal accumulation attenuation coefficient and the thermal accumulation oscillation coupling coefficient to complete the prediction calculation of active and reactive power adjustable capabilities.

6. The method for predicting the adjustable capacity of distributed energy resources based on multimodal data fusion according to claim 1, characterized in that, The corrected micro-oscillation angular frequency is obtained by converting the oscillation frequency in the corrected power grid oscillation parameters. The corrected micro-oscillation amplitude is consistent with the oscillation amplitude in the corrected power grid oscillation parameters. The micro-oscillation attenuation coefficient is obtained by performing frequency domain decomposition on the corrected power grid oscillation parameters.

7. The method for predicting the adjustable capacity of distributed energy resources based on multimodal data fusion according to claim 1, characterized in that, The calculation process of the spatiotemporal heterogeneous attenuation correction formula for multimodal data is as follows: Substitute the spatial attenuation factor, temporal attenuation factor, heterogeneous weight of data type, heterogeneous coefficient of sampling frequency, and the original single-mode parameter values ​​corresponding in the original multimodal dataset into the spatiotemporal heterogeneous attenuation correction formula for multimodal data. Perform calculations on electrical parameters, equipment thermal parameters, and power grid oscillation parameters respectively to obtain the corrected electrical parameters, corrected equipment thermal parameters, and corrected power grid oscillation parameters. Integrate the above corrected parameters to obtain the corrected basic operating parameter set.

8. The method for predicting the adjustable capacity of distributed energy sources based on multimodal data fusion according to claim 5, characterized in that, The calculation process of the equipment thermal cumulative attenuation coefficient is as follows: extract the corrected device heat flux density and corrected micro-oscillation amplitude from the corrected basic operating parameter set; obtain the corrected device temperature rise by calculating the difference between the actual device temperature and the ambient temperature in the corrected device thermal parameters; substitute the corrected device heat flux density, corrected device temperature rise, corrected micro-oscillation amplitude, device heat dissipation area, thermal accumulation time, device material specific heat capacity, device heating core mass, power device thermal resistance coefficient, power grid micro-oscillation influence coefficient, and power grid micro-oscillation reference amplitude into the equipment thermal cumulative attenuation coefficient calculation formula, and perform fractional operation, multiplication operation, and subtraction operation in sequence to obtain the thermal cumulative attenuation coefficient.

9. The method for predicting the adjustable capacity of distributed energy resources based on multimodal data fusion according to claim 5, characterized in that, The calculation process of the coupled adjustable capacity comprehensive prediction formula is as follows: First, the thermal cumulative oscillation coupling coefficient is obtained through the thermal cumulative oscillation coupling coefficient calculation formula. Then, the corrected active power, corrected reactive power, and corrected micro-oscillation angular frequency are extracted from the corrected basic operating parameter set. The corrected active power, thermal cumulative attenuation coefficient, thermal cumulative oscillation coupling coefficient, corrected micro-oscillation angular frequency, prediction time window, and distributed energy regulation response time are substituted into the active power adjustable capacity prediction formula to complete the calculation and obtain the active power adjustable capacity. At the same time, the corrected reactive power, thermal cumulative attenuation coefficient, thermal cumulative oscillation coupling coefficient, corrected micro-oscillation angular frequency, prediction time window, and distributed energy regulation response time are substituted into the reactive power adjustable capacity prediction formula to complete the calculation and obtain the reactive power adjustable capacity. The prediction time window is the prediction duration of the distributed energy adjustable capacity, and the distributed energy regulation response time is the power regulation response duration of the distributed energy after receiving the dispatch command.

Images