Analysis Method and System for Carbon Electricity Storage Power Control Strategy Considering Multi-Source Load Coordination

Abstract

This application relates to the field of power control strategy analysis technology, and discloses a method and system for analyzing carbon-electric energy storage power control strategies considering multi-source-load coordination. The method includes: performing power balance analysis on the source-load power coordination feature matrix under carbon emission constraints to obtain an initial solution for power scheduling of the carbon-electric energy storage system; performing source-load coordination optimization on the initial solution to obtain a power optimization sequence; performing energy storage capacity allocation on the power optimization sequence to obtain a charging and discharging power command set; and performing power correction on the charging and discharging power command set under carbon emission cost constraints to obtain an energy storage scheduling strategy, which is then subjected to multi-objective coordination optimization to obtain a carbon-electric energy storage power control strategy. This application, considering carbon emission constraints, achieves accurate extraction of source-load power characteristics and multi-objective coordination optimization, thereby improving the power control performance of the carbon-electric energy storage system.

Description

Technical Field

[0001] This application relates to the field of power control strategy analysis, and in particular to a method and system for analyzing carbon-electric energy storage power control strategies that consider multi-source load coordination. Background Technology

[0002] With the large-scale integration of renewable energy, carbon-based energy storage systems are playing an increasingly important role in power systems. Traditional power systems primarily employ centralized control strategies, achieving stable operation through coordinated control of the generation, storage, and load sides. Currently, domestic and international scholars have conducted extensive research on power control strategies for carbon-based energy storage systems, including predictive control-based energy storage power optimization methods, distributed control strategies considering load response, and multi-objective optimization-based collaborative scheduling methods. These methods have improved the operating efficiency of carbon-based energy storage systems to some extent.

[0003] However, existing control strategies generally suffer from the following shortcomings: First, the analysis of source-load power characteristics is not in-depth enough, making it difficult to accurately grasp the complementary relationship between source and load; second, when considering carbon emission constraints, static constraints are often used, which cannot adapt to the dynamic changes in carbon emission costs; third, the coordination between energy storage capacity allocation and charge / discharge power control is insufficient, resulting in low overall system operating efficiency; and finally, the trade-offs between multiple objectives are relatively simple, making it difficult to achieve balanced optimization of each objective. Summary of the Invention

[0004] This application provides a method and system for analyzing carbon-electric energy storage power control strategies that consider multi-source load coordination. Under the condition of considering carbon emission constraints, it can achieve accurate extraction of source load power characteristics and multi-objective coordinated optimization, thereby improving the power control performance of carbon-electric energy storage systems.

[0005] Firstly, this application provides an analysis method for carbon-based electric energy storage power control strategies considering multi-source load synergy, the analysis method for carbon-based electric energy storage power control strategies considering multi-source load synergy includes:

[0006] Multi-time-series feature extraction processing is performed on the source-load power data of the carbon-electric energy storage system to obtain the source-load power collaborative feature matrix.

[0007] The power balance analysis under carbon emission constraints is performed on the source-load power coordination characteristic matrix to obtain the initial solution of power scheduling for the carbon-electric energy storage system.

[0008] The initial power scheduling solution is subjected to source-load cooperative optimization processing to obtain a power optimization sequence that considers the complementary characteristics of multiple source loads;

[0009] The power optimization sequence is processed for energy storage capacity allocation to obtain the charging and discharging power instruction set of the energy storage unit;

[0010] The charging and discharging power instruction set is subjected to power correction processing under carbon emission cost constraints to obtain a carbon-electricity linked energy storage scheduling strategy.

[0011] The carbon-electricity linkage energy storage scheduling strategy is subjected to multi-objective collaborative optimization to obtain a carbon-electricity energy storage power control strategy.

[0012] Secondly, this application provides a carbon-electric energy storage power control strategy analysis system considering multi-source load synergy, the carbon-electric energy storage power control strategy analysis system considering multi-source load synergy includes:

[0013] The acquisition module is used to perform multi-time-series feature extraction processing on the source-load power data of the carbon-electric energy storage system to obtain the source-load power collaborative feature matrix.

[0014] The processing module is used to perform power balance analysis on the source-load power coordination feature matrix under carbon emission constraints to obtain the initial solution for power scheduling of the carbon-electric energy storage system.

[0015] The optimization module is used to perform source-load collaborative optimization processing on the initial power scheduling solution to obtain a power optimization sequence that considers the complementary characteristics of multiple source loads.

[0016] The allocation module is used to perform energy storage capacity allocation processing on the power optimization sequence to obtain the charging and discharging power instruction set of the energy storage unit;

[0017] The correction module is used to perform power correction processing on the charging and discharging power instruction set under the constraint of carbon emission cost, so as to obtain a carbon-electricity linked energy storage scheduling strategy.

[0018] The coordination module is used to perform multi-objective collaborative optimization processing on the carbon-electric linkage energy storage scheduling strategy to obtain the carbon-electric energy storage power control strategy.

[0019] The technical solution provided in this application, by performing multi-time-series feature extraction processing on the source-load power data of the carbon-electric energy storage system, can comprehensively and accurately obtain the dynamic variation characteristics of source-load power, providing a reliable data foundation for subsequent collaborative optimization. Simultaneously, by performing power balance analysis under carbon emission constraints on the source-load power collaborative feature matrix, the power balance of the system can be ensured while meeting carbon emission requirements, avoiding system instability caused by power imbalance. Furthermore, the source-load collaborative optimization processing of the initial power scheduling solution fully considers the complementary characteristics between sources and loads, maximizing the advantages of various sources and loads and improving the overall operating efficiency of the system. In addition, by performing energy storage capacity allocation processing on the power optimization sequence, it is possible to... By rationally allocating energy storage resources and avoiding overuse or waste of energy storage capacity, the efficient utilization of energy storage units can be ensured. Furthermore, power correction processing under carbon emission cost constraints is applied to the charging and discharging power command set, which can dynamically adjust the charging and discharging strategy to adapt to changes in carbon emission costs, effectively reducing the system's operating costs. Finally, through multi-objective collaborative optimization of the carbon-electric linkage energy storage scheduling strategy, a balanced optimization of multiple objectives such as economy, environmental protection, and reliability is achieved. This makes the final carbon-electric energy storage power control strategy highly practical and adaptable, ensuring not only the stable operation of the system but also maximizing economic benefits while reducing carbon emissions, significantly improving the overall performance and control effect of the carbon-electric energy storage system. Attached Figure Description

[0020] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0021] Figure 1 This is a schematic diagram of an embodiment of the carbon electric energy storage power control strategy analysis method considering multi-source load coordination in this application.

[0022] Figure 2 This is a schematic diagram of an embodiment of the carbon-electric energy storage power control strategy analysis system that considers multi-source load coordination in the embodiments of this application. Detailed Implementation

[0023] This application provides a method and system for analyzing carbon-electric energy storage power control strategies that consider multi-source load synergy. The terms "first," "second," "third," "fourth," etc. (if present) in the specification, claims, and accompanying drawings of this application are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments described herein can be implemented in a sequence other than that illustrated or described herein. Furthermore, the terms "comprising" or "having" and any variations thereof are intended to cover a non-exclusive inclusion; for example, a process, method, system, product, or device that includes a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or devices.

[0024] For ease of understanding, the specific process of the embodiments of this application is described below. Please refer to [link / reference]. Figure 1 One embodiment of the carbon-electric energy storage power control strategy analysis method considering multi-source load synergy in this application includes:

[0025] Step S101: Perform multi-time-series feature extraction processing on the source-load power data of the carbon-electric energy storage system to obtain the source-load power collaborative feature matrix;

[0026] Step S102: Perform power balance analysis on the source-load power cooperative characteristic matrix under carbon emission constraints to obtain the initial solution for power scheduling of the carbon-electric energy storage system;

[0027] Step S103: Perform source-load collaborative optimization on the initial power scheduling solution to obtain a power optimization sequence that considers the complementary characteristics of multiple source loads;

[0028] Step S104: Perform energy storage capacity allocation processing on the power optimization sequence to obtain the charging and discharging power instruction set of the energy storage unit;

[0029] Step S105: Perform power correction processing on the charging and discharging power instruction set under the constraint of carbon emission cost to obtain the carbon-electricity linkage energy storage scheduling strategy.

[0030] Step S106: Perform multi-objective collaborative optimization on the carbon-electricity linkage energy storage scheduling strategy to obtain the carbon-electricity energy storage power control strategy.

[0031] It is understood that the executing entity of this application can be a carbon-electric energy storage power control strategy analysis system considering multi-source load coordination, or it can be a terminal or a server; the specific implementation is not limited here. This application's embodiments use a server as an example for illustration.

[0032] Specifically, multi-time-series feature extraction processing is performed on source-load power data. Historical power data is divided into time windows, such as dividing a day into 96 time points, with a sampling point every 15 minutes. For a month's data, this results in 2880 time-series data points. After standardization, wavelet transform is used to decompose the power data into different frequency components, thereby identifying the main characteristics of power fluctuations. For example, wavelet transform can decompose photovoltaic power generation data into three levels: long-term trend, intraday periodic variation, and short-term random fluctuation. Fourier transform is then used to analyze the periodic characteristics of these components, obtaining a power fluctuation feature sequence. After obtaining the power fluctuation feature sequence, time-series correlation analysis is used to determine the correlation between different source loads, and principal component analysis is used to reduce data dimensionality. Specifically, by calculating the correlation coefficient matrix, the main feature vectors are extracted. In one practical case, the original 100-dimensional power feature data was reduced to 20 dimensions through principal component analysis, while retaining more than 85% of the information. Then, the K-means clustering algorithm was used to classify the dimensionality-reduced features, dividing the power features into different categories such as typical workdays and rest days, and reconstructing the power sequence based on these categories to finally obtain the source-load power collaborative feature matrix.

[0033] For power balance analysis under carbon emission constraints, a carbon emission factor mapping relationship is first established to convert the power values ​​of different types of power sources and loads into corresponding carbon emissions. By performing time-series decomposition on the carbon emission characteristic data, a baseline carbon emission sequence is obtained, and constraint conditions are extracted to form a constraint boundary set. In practical applications, the baseline carbon emission value for a certain regional power grid is 500g. The power consumption per kWh is converted into specific power constraints through dynamic threshold calculation, thereby constructing power balance constraints and achieving initial allocation of source and load power. In the source-load co-optimization stage, the complementary relationships between different source and load types are analyzed to construct a complementarity matrix. For example, in a certain region, the complementarity between photovoltaic power generation and industrial load is 0.75, and the complementarity between wind power and commercial load is 0.68. Based on these complementary characteristics, the matching degree index of source and load power is calculated, generating corresponding power regulation coefficients. An adaptive weight allocation algorithm is used to calculate the weights of different objectives, obtaining optimized weight coefficients, which are then used to adjust the power dispatch scheme, forming a source-load co-optimization candidate sequence.

[0034] In the energy storage capacity allocation stage, demand analysis is performed on the power optimization sequence. A day is divided into 96 time periods, and the energy storage capacity demand for each period is calculated considering factors such as charge / discharge efficiency and state of charge. For example, if the energy storage capacity demand for a certain period is 2MWh, considering a charge / discharge efficiency of 0.9, the actual allocated energy storage capacity is 2.22MWh. Constraint verification ensures that the allocation scheme meets the technical limitations of the energy storage system, ultimately generating a charge / discharge power instruction set. For power correction under carbon emission cost constraints, the carbon emission cost for different time periods is first calculated. For example, when the carbon trading price is 50 yuan / ton, the carbon emission cost sequence is calculated based on the power data. Based on the cost constraints, the power correction amount is calculated, and the carbon-electricity linkage characteristics after correction are analyzed. Through multiple rounds of iterative optimization, a scheduling strategy that meets carbon emission constraints and has good economic efficiency is generated.

[0035] In the multi-objective collaborative optimization stage, the carbon-electricity linkage energy storage dispatch strategy is decomposed into multiple objectives such as economy, environmental protection, and reliability. The importance of each objective is determined by calculating objective weights. For example, the weights for economy, environmental protection, and reliability are 0.4, 0.3, and 0.3, respectively. Based on these weights, the strategy is adjusted, and the optimal control strategy is selected through comprehensive performance evaluation.

[0036] Taking a microgrid in a certain region as an example, this region includes 3MW of photovoltaic power generation, 2MW of wind power generation, 4MWh of energy storage system, and 5MW of electricity load. Data from 96 time periods throughout the day was processed using the above method. First, the source-load power characteristics were extracted, revealing a significant inverse correlation between photovoltaic power generation and air conditioning load, with a correlation coefficient of -0.82. Under a carbon emission constraint of 400g... With a capacity of / kWh, an initial dispatch scheme was obtained through power balance analysis. Considering the complementary characteristics of source and load, the energy storage capacity was allocated at different times, such as allocating 1.5MWh for energy storage during peak photovoltaic power generation and 2MWh for discharge during peak electricity consumption. The final control strategy reduced the system's carbon emission intensity to 385g. / kWh, while ensuring a renewable energy consumption rate of over 95%, and reducing the overall operating cost of the system by 15% compared to traditional dispatching methods.

[0037] In this embodiment, by performing multi-time-series feature extraction processing on the source-load power data of the carbon-electric energy storage system, the dynamic variation characteristics of source-load power can be comprehensively and accurately obtained, providing a reliable data foundation for subsequent collaborative optimization. Simultaneously, by performing power balance analysis under carbon emission constraints on the source-load power collaborative feature matrix, the power balance of the system can be ensured while meeting carbon emission requirements, avoiding system instability caused by power imbalance. Furthermore, performing source-load collaborative optimization processing on the initial power scheduling solution fully considers the complementary characteristics between sources and loads, maximizing the advantages of various sources and loads and improving the overall operating efficiency of the system. In addition, by performing energy storage capacity allocation processing on the power optimization sequence, reasonable... By allocating energy storage resources to avoid overuse or waste of energy storage capacity and ensuring efficient utilization of energy storage units, and further by performing power correction processing on the charging and discharging power command set under carbon emission cost constraints, the charging and discharging strategy can be dynamically adjusted to adapt to changes in carbon emission costs, effectively reducing the system's operating costs. Finally, through multi-objective collaborative optimization of the carbon-electric linkage energy storage scheduling strategy, a balanced optimization of multiple objectives such as economy, environmental protection, and reliability is achieved. This makes the final carbon-electric energy storage power control strategy highly practical and adaptable, ensuring not only stable system operation but also maximizing economic benefits while reducing carbon emissions, significantly improving the overall performance and control effect of the carbon-electric energy storage system.

[0038] In one specific embodiment, the process of performing step S101 may specifically include the following steps:

[0039] (1) The historical power data of the carbon-electric energy storage system is divided into time windows to obtain a power data sequence of length T. The power data sequence is then standardized to obtain standardized power data, where T is a positive integer greater than 1.

[0040] (2) Wavelet transform decomposition is performed on the standardized power data to obtain the multi-scale time-frequency features of the power data, and the periodic analysis of the multi-scale time-frequency features is performed by Fourier transform to obtain the power fluctuation feature sequence.

[0041] (3) Perform time-series correlation analysis on the power fluctuation feature sequence to obtain the correlation feature vector of source load power, and perform principal component analysis on the correlation feature vector to obtain the feature set after dimensionality reduction.

[0042] (4) The power feature clustering process is performed on the dimensionality-reduced feature set by K-means clustering algorithm to obtain the typical feature categories of source load power, and the power data is reconstructed according to the typical feature categories to obtain the reconstructed power sequence.

[0043] (5) Perform time-series feature encoding on the reconstructed power sequence to obtain the time-series feature encoding matrix of source-load power, and construct the source-load power collaborative feature matrix based on the time-series feature encoding matrix.

[0044] Specifically, time windows are divided by selecting an appropriate time window length T to segment the data. The selection of the time window is based on the data sampling frequency and analysis requirements. For example, for power data sampled every 15 minutes, there are 96 data points per day. If a 7-day time window is selected, then T=672. The segmented power data sequence is then standardized to convert the data to a standard normal distribution with a mean of 0 and a variance of 1, eliminating the influence of different dimensions. Subsequently, wavelet transform is used to perform multi-scale decomposition on the standardized power data. Wavelet transform can perform localized analysis of the signal in both the time and frequency domains simultaneously. By selecting an appropriate wavelet basis function (such as the db4 wavelet), the power data is decomposed into three levels to obtain the time-frequency characteristics of different frequency components. For example, for photovoltaic power generation data, the first level of decomposition yields low-frequency components (0-0.5Hz) and high-frequency components (0.5-1Hz). The second level further decomposes the low-frequency components into components of 0-0.25Hz and 0.25-0.5Hz. The third level continues to decompose the lowest frequency components. These multi-scale time-frequency characteristics are periodically analyzed using Fourier transform to calculate the amplitude and phase characteristics of different frequency components, forming a power fluctuation characteristic sequence.

[0045] For the obtained power fluctuation characteristic sequence, time-series correlation analysis is performed to calculate the correlation coefficients between different source loads and construct a correlation feature vector. For example, in a carbon-electric energy storage system in a certain region, there are four types of power generation: photovoltaic power generation, wind power generation, industrial load, and commercial load. By calculating the Pearson correlation coefficients between them, a 6-dimensional correlation feature vector is obtained. Principal component analysis is performed on this feature vector. By calculating eigenvalues ​​and eigenvectors, principal components with a cumulative contribution rate of over 85% are selected as the feature set after dimensionality reduction. The K-means clustering algorithm is used to perform cluster analysis on the dimensionality-reduced feature set. K-means clustering divides the data points into K clusters through iterative optimization. First, K cluster centers are randomly selected, and then the following steps are repeated: calculate the distance from each data point to each cluster center, assign the data point to the cluster to which the nearest cluster center belongs, recalculate the center point of each cluster, until the cluster centers no longer change significantly. The optimal K value is determined by calculating the silhouette coefficient, and the typical feature categories of source load power are obtained. Based on these categories, the time-series features of the original power data are reconstructed to form a reconstructed power sequence.

[0046] The reconstructed power sequence is encoded using temporal features. A sliding time window method is used to extract local temporal features, and one-hot encoding is employed to convert categorical features into numerical features, constructing a temporal feature encoding matrix. Finally, the temporal feature encoding matrix is ​​combined with the correlation features of source and load power to generate a source-load power co-factor feature matrix.

[0047] Taking a carbon-electric energy storage system in a certain region as an example, this system includes 2MW of photovoltaic power generation, 1.5MW of wind power generation, 3MWh of energy storage devices, and 4MW of integrated load. One month's historical power data (sampling interval of 15 minutes) was processed. First, the data was divided into 7-day time windows, resulting in 4 time window data sequences. The data for each time window was standardized, and then the standardized data was decomposed into 3 levels using db4 wavelets to obtain the time-frequency characteristics of 7 frequency bands. Fourier transform analysis revealed that photovoltaic power generation exhibits significant periodicity in the 24-hour and 12-hour periods, with amplitudes of 1.2 and 0.8, respectively; wind power shows strong fluctuation characteristics in the 6-hour and 3-hour periods, with amplitudes of 0.9 and 0.6, respectively. In time-series correlation analysis, the correlation coefficients between photovoltaic power generation and load were calculated to be -0.75, wind power and load to be -0.45, and photovoltaic and wind power to be 0.25. Principal component analysis was used to select the first two principal components (cumulative contribution rate 87%) as the features after dimensionality reduction. K-means clustering was then used to cluster these features. By calculating the silhouette coefficients for different K values ​​(K=2~8), the optimal K value was determined to be 4, resulting in four typical power characteristics: high photovoltaic power with low load, high wind power with medium load, low power generation with high load, and balanced. These four types of features were then temporally encoded, with each type represented by a 4-dimensional vector, ultimately constructing a 96×16 source-load power collaborative feature matrix, where 96 represents the number of time points in a day, and 16 represents the feature dimension at each time point.

[0048] In one specific embodiment, the process of performing step S102 may specifically include the following steps:

[0049] (1) Carbon emission factor mapping is performed on the source-load power cooperative feature matrix to obtain carbon emission feature dataset, and time series decomposition is performed on the carbon emission feature dataset to obtain carbon emission benchmark sequence;

[0050] (2) Extract the constraints from the carbon emission baseline sequence to obtain the carbon emission constraint boundary set, and perform dynamic threshold calculation on the carbon emission constraint boundary set to obtain the carbon emission constraint threshold.

[0051] (3) The carbon emission constraint threshold and power data are processed to form a balance constraint, and the power balance constraint conditions are obtained. The power of the source load is initially allocated according to the power balance constraint conditions to obtain the power allocation scheme.

[0052] (4) Perform power balance verification processing according to the power allocation scheme to obtain a set of feasible power scheduling solutions, and perform screening processing on the set of feasible power scheduling solutions to obtain the initial power scheduling solution of the carbon-electric energy storage system.

[0053] Specifically, corresponding carbon emission factors are assigned to different types of power sources and loads: the carbon emission factor for thermal power units is 920g. / kWh, the carbon emission factor for photovoltaic and wind power generation is 0g / kWh, and the indirect carbon emission factor for industrial load is 650g. / kWh, the indirect carbon emission factor for commercial load is 450g. / kWh. These carbon emission factors are multiplied by the corresponding power values ​​in the source-load power co-feature matrix to obtain a dataset reflecting the carbon emission characteristics of each time period. The carbon emission characteristic dataset is then subjected to time-series decomposition, using a seasonal decomposition method to divide the data into three components: a trend component, a seasonal component, and a random component. The trend component reflects the long-term trend of carbon emissions, the seasonal component reflects the intraday and weekly periodic variations, and the random component represents short-term fluctuations. By calculating the statistical characteristics of each component, such as the mean, standard deviation, and coefficient of variation, a carbon emission baseline series is formed.

[0054] During the constraint extraction process, the upper and lower boundaries of carbon emission constraints are extracted based on the carbon emission baseline sequence. The upper boundary is determined by the carbon emission limits stipulated by the country or region, such as a carbon emission intensity limit of 550g for a certain region. / kWh; the lower bound considers technical and economic feasibility, such as taking into account the instability of photovoltaic power generation, setting the lower limit of carbon emission intensity at 200g. / kWh. Dynamic threshold calculations are performed on these constraint boundaries, dynamically adjusting carbon emission constraint thresholds based on factors such as current load levels, renewable energy output forecasts, and carbon trading prices. When constructing balance constraints, carbon emission constraint thresholds are converted into power balance constraints. Based on the power balance principle, the total power generated at any given time must equal the sum of the electricity load and the energy storage charging / discharging power. Simultaneously, considering the carbon emission constraint thresholds, the upper and lower limits of output for different power sources are calculated. For energy storage devices, technical constraints such as state of charge and charging / discharging efficiency need to be considered, such as a charging / discharging efficiency of 90% and a maximum charging / discharging power of 80% of the rated power. Based on these constraints, a linear programming method is used to initially allocate source and load power, obtaining power allocation schemes for each time period. Power balance verification is performed based on the power allocation schemes to check whether the power balance for each time period meets the constraints, including power balance constraints, carbon emission constraints, and equipment operation constraints. Schemes that meet all constraints are included in the feasible solution set for power scheduling. Then, the schemes in the feasible solution set are comprehensively evaluated and screened. The evaluation indicators include carbon emissions, economic costs and system stability. Finally, the scheme with the best overall performance is selected as the initial solution for power scheduling of the carbon-electric energy storage system.

[0055] Taking a carbon-electric energy storage system in an industrial park as an example, the system includes 3MW of photovoltaic power generation, 2MW of wind power generation, a 5MWh energy storage system, and a 6MW integrated load. Carbon emission factor mapping was performed on the source-load power coordination characteristic matrix to obtain a carbon emission characteristic dataset reflecting 96 time points over 24 hours. Time-series decomposition revealed that carbon emissions generally exhibit a cyclical variation, being higher on weekdays and lower on weekends, with a daily peak-to-valley difference of 300g. / kWh. In accordance with the park's carbon emission management requirements, the carbon emission intensity is set at a maximum of 500g. / kWh, lower limit is 150g / kWh. At 9:00 AM, the load demand is 4MW, the predicted output of photovoltaic (PV) is 2MW, the predicted output of wind power is 1MW, and the carbon trading price is 45 yuan / ton. Based on dynamic threshold calculations, the carbon emission constraint for this period is 400gCO2 / kWh. Considering that the actual output fluctuation of PV power generation is within ±10% of the predicted value, and the fluctuation of wind power is within ±15%, and the current state of charge of the energy storage system is 60%, the calculated power allocation scheme for this period is: actual PV output 1.8MW, actual wind power output 0.9MW, and energy storage discharge 0.8MW, taking into account a 90% charge / discharge efficiency. After power balance verification, this scheme meets all operational constraints, and the average carbon emission intensity for the period is 385gCO2 / kWh. / kWh was included in the feasible solution set. Ultimately, through comprehensive evaluation, this scheme was selected as the initial power scheduling solution for this time period.

[0056] In one specific embodiment, the process of executing step S103 may specifically include the following steps:

[0057] (1) The source-load complementarity characteristic analysis is performed on the initial solution of power scheduling to obtain the source-load complementarity relationship matrix, and the source-load complementarity relationship matrix is ​​segmented into time series to obtain the time-segmented complementary feature set;

[0058] (2) The source-load power matching degree is calculated and processed for the complementary feature set of different time periods to obtain the source-load power matching degree index. The power adjustment coefficient is generated based on the source-load power matching degree index to obtain the power adjustment coefficient sequence.

[0059] (3) Perform source-load co-optimization on the power regulation coefficient sequence to obtain the co-optimization objective function, and perform weight calculation on the co-optimization objective function through an adaptive weight allocation algorithm to obtain the optimized weight coefficients;

[0060] (4) The power scheduling scheme is adjusted in a coordinated manner according to the optimized weight coefficients to obtain the source-load coordinated candidate sequence, and the complementary performance evaluation of the source-load coordinated candidate sequence is performed to obtain the complementary performance index set;

[0061] (5) Perform comprehensive evaluation on the complementary performance index set to obtain the source-load synergistic optimization results, and construct a power optimization sequence that considers the complementary characteristics of multiple source loads based on the source-load synergistic optimization results.

[0062] Specifically, the initial solution for power scheduling is analyzed for source-load complementarity characteristics. A source-load complementarity matrix is ​​constructed by calculating the power complementarity index between different source loads. The source-load power complementarity index represents the degree of inverse correlation between power fluctuations of different types of source loads, with a value range of [-1, 1]. The closer the value is to -1, the stronger the complementarity. For each pair of source loads, the inverse of the Pearson correlation coefficient is calculated based on its power time series data as the complementarity index, thus forming an n×n complementarity matrix, where n is the total number of source load types. The source-load complementarity matrix is ​​then segmented into time series segments, dividing the 24 hours into typical time periods, such as dividing a day into the early morning period (0:00-6:00), the daytime period (6:00-18:00), and the nighttime period (18:00-24:00). Within each time period, complementary features between source loads are extracted, including indicators such as power complementarity strength, complementarity duration, and complementarity stability, forming a time-segmented complementarity feature set. Normalize each feature in the complementary feature set so that its value is uniformly within the range of [0,1].

[0063] The source-load power matching degree index is calculated based on time-segmented complementary feature sets. This index reflects the degree of matching between source and load in terms of power magnitude and temporal characteristics. The matching degree index consists of two parts: power capacity matching degree and temporal characteristic matching degree. The former considers the matching relationship of source-load power amplitude, while the latter focuses on the consistency of power change trends. Based on the power matching degree index, a corresponding power adjustment coefficient sequence is generated to adjust the power output of source and load. The adjustment coefficient ranges from [0.8, 1.2], where a value greater than 1 indicates that power needs to be increased, and a value less than 1 indicates that power needs to be reduced. Source-load collaborative optimization is performed on the power adjustment coefficient sequence to construct a multi-objective optimization function that includes economic, environmental, and reliability objectives. The economic objective considers operating costs and carbon trading costs, the environmental objective focuses on total carbon emissions, and the reliability objective reflects the stability of power balance. An adaptive weight allocation algorithm dynamically adjusts the weight coefficients of each objective. This algorithm updates the weight allocation ratio in real time based on the gradient information of the objective function value, enabling the weight coefficients to adapt to changes in the objectives during the optimization process.

[0064] The power scheduling scheme is adjusted based on the obtained optimization weight coefficients, generating multiple source-load collaborative candidate sequences. The complementary performance of each candidate sequence is evaluated, calculating performance indicators including power complementarity, fluctuation suppression rate, and load following error. Power complementarity measures the complementary effect between source and load, fluctuation suppression rate reflects the ability to mitigate power fluctuations, and load following error represents the accuracy of response to load changes. Finally, the set of complementary performance indicators is comprehensively evaluated. The analytic hierarchy process (AHP) is used to determine the importance of each indicator, calculate the comprehensive score, and select the scheme with the highest score as the source-load collaborative optimization result. Based on this, a power optimization sequence considering the complementary characteristics of multiple source and load is constructed.

[0065] Taking a carbon-electric energy storage system in an industrial park as a specific case, this system includes 2.5MW of photovoltaic power generation, 2MW of wind power generation, a 4MWh energy storage system, and 5MW of industrial load. First, the source-load complementarity matrix is ​​calculated. The complementarity index between photovoltaic power generation and industrial load is -0.82, the complementarity index between wind power and industrial load is -0.58, and the complementarity index between photovoltaic and wind power is 0.35. After dividing the day into three time periods, from 6:00 to 18:00, the power output of photovoltaic power generation ranges from 0.5 to 2.3MW, and the power output of industrial load ranges from 2.8 to 4.5MW. The calculated power capacity matching degree is 0.78, and the time-series characteristic matching degree is 0.85. Based on these matching degree indices, power regulation coefficients are generated. For example, at 10:00, the regulation coefficient for photovoltaic power generation is 1.15 (requiring increased generation), and the regulation coefficient for wind power generation is 0.92 (requiring reduced generation). The initial weights for economic efficiency, environmental friendliness, and reliability objectives were calculated using an adaptive weight allocation algorithm, with initial weights of 0.4, 0.35, and 0.25, respectively. During the optimization process, due to increased carbon emission pressure, the weight of the environmental friendliness objective was dynamically adjusted to 0.42, while the weights of the economic efficiency and reliability objectives were adjusted accordingly to 0.35 and 0.23. Based on the adjusted weights, source-load power was collaboratively optimized, generating five candidate schemes. After complementary performance evaluation, the optimal scheme at 14:00 had the following data: actual photovoltaic output of 2.1MW, actual wind power output of 1.6MW, energy storage charging of 0.5MW, and industrial load of 3.2MW. This scheme achieved a power complementarity of 0.88, a fluctuation suppression rate of 82%, and a load following error of less than 5%.

[0066] In one specific embodiment, the process of executing step S104 may specifically include the following steps:

[0067] (1) Perform energy storage capacity demand analysis on the power optimization sequence to obtain the energy storage capacity demand feature set, and perform time period division on the energy storage capacity demand feature set to obtain the time period capacity demand sequence.

[0068] (2) Perform energy storage unit status assessment processing on the time-segmented capacity demand sequence to obtain the energy storage status feature matrix, and perform charge and discharge efficiency calculation processing on the energy storage status feature matrix to obtain the charge and discharge efficiency coefficient.

[0069] (3) The energy storage capacity is initially allocated based on the charge and discharge efficiency coefficient to obtain the energy storage capacity allocation scheme, and the energy storage capacity allocation scheme is constrained and verified to obtain the feasible capacity allocation set;

[0070] (4) Perform charging and discharging power calculation on the feasible capacity allocation set to obtain the charging and discharging power sequence, and perform power command generation processing based on the charging and discharging power sequence to obtain the charging and discharging power command set of the energy storage unit.

[0071] Specifically, energy storage capacity demand analysis is performed on the power optimization sequence, and the energy storage capacity demand characteristic set is obtained by calculating the power fluctuation characteristics. The specific analysis process is based on the following equation:

[0072]

[0073] In the formula, This represents the energy storage capacity allocation value at time t. Let be the weight coefficient of the i-th allocation factor, with a value range of [0,1]. This is the power balance factor, used to balance real-time power deviation. The power change rate coefficient reflects the dynamic characteristics of power change. The historical power integral coefficient takes into account the historical power accumulation effect. This is the SOC deviation coefficient, reflecting the state of charge constraint. This is the net power value. This refers to the state-of-charge deviation. In practical applications, such as the energy storage capacity allocation in an industrial park, when the power fluctuation is ±2MW, it is calculated to obtain... =0.6, =0.3, =0.1, which effectively suppresses power fluctuations.

[0074] Furthermore, the energy storage capacity demand characteristic set is divided into time periods, with the 24-hour period divided into peak, valley, and flat periods according to load characteristics. Within each time period, the correlation between characteristics is calculated:

[0075]

[0076] In the formula, The total correlation of features, The association weights for feature pairs (i,j) The feature covariance represents the correlation between features. The characteristic standard deviation, This is the correlation function for capacity change. This is the state-of-charge (SOC) correlation function. Analysis of a photovoltaic power generation and energy storage project shows that the characteristic correlation degrees for peak and valley periods reach 0.85 and 0.72, respectively, based on which the time-of-use capacity demand sequence was determined. The energy storage unit state was assessed based on the time-of-use capacity demand sequence, constructing a state characteristic matrix including SOC, cycle life, and temperature characteristics. The assessment process used a weighted scoring method, limiting the SOC to 15%-95%, the cycle life to 6000 complete cycles, and the temperature to -10℃ to 45℃. Based on the state characteristic matrix, the charge and discharge efficiencies were calculated. Under normal operating conditions, the charging efficiency is 92% and the discharging efficiency is 95%, forming an efficiency coefficient matrix.

[0077] The initial allocation of energy storage capacity is based on the charge / discharge efficiency coefficient, prioritizing allocation to periods and conditions with higher efficiency. Taking a 4MWh energy storage device as an example, with an efficiency coefficient of 0.92, 1.8MWh is allocated during peak hours, 1.2MWh during normal hours, and 1MWh during off-peak hours. The allocation scheme is constrained and verified to ensure compliance with power limits (≤1.2 times rated power), SOC range (15%-95%), and cycle count limits (≤2 cycles per day). Finally, charge / discharge power calculations are performed on the feasible capacity allocation set, generating power commands based on real-time power balance principles. For example, during a certain period, when the photovoltaic power is 2.5MW and the load demand is 3MW, the calculated energy storage discharge power is 0.5MW, lasting 30 minutes. The power command generation considers constraints such as ramp-up limits (≤20% of rated power / minute) and minimum start / stop time (≥15 minutes). In a comprehensive case study, an industrial park configured a 5MWh energy storage device, and data from 96 time points throughout the day were analyzed using the above method. First, the capacity demand characteristics were calculated, resulting in a maximum power fluctuation of 2.8MW lasting 45 minutes. After dividing the time periods, the peak demand (10:00-15:00) was 2.2MWh, the normal demand (7:00-10:00, 15:00-18:00) was 1.6MWh, and the off-peak demand (the remaining time) was 1.2MWh. Status assessment showed a current SOC of 65%, having completed 1200 charge-discharge cycles with a charging efficiency of 0.91 and a discharging efficiency of 0.94. Based on this, the generated charge / discharge power command was set at 14:00 for a charging power of 1.5MW (lasting 20 minutes). Considering a ramp rate limit of 15% / minute, efficient utilization of the energy storage capacity was achieved.

[0078] In one specific embodiment, the process of executing step S105 may specifically include the following steps:

[0079] (1) Perform carbon emission cost calculation on the charge and discharge power instruction set to obtain the carbon emission cost sequence, and perform threshold constraint analysis on the carbon emission cost sequence to obtain the cost constraint conditions;

[0080] (2) The power correction amount is calculated based on the cost constraint to obtain the power correction amount sequence, and the power correction amount sequence is subjected to carbon-electric linkage analysis to obtain the carbon-electric linkage feature set.

[0081] (3) Perform scheduling strategy generation processing on the carbon-electric linkage feature set to obtain the carbon-electric linkage scheduling scheme, and perform feasibility verification processing on the carbon-electric linkage scheduling scheme to obtain the feasible scheduling scheme set.

[0082] (4) Perform scheduling performance evaluation on the set of feasible scheduling schemes to obtain scheduling performance indicators, and then screen the scheduling schemes according to the scheduling performance indicators to obtain the carbon-electricity linkage energy storage scheduling strategy.

[0083] Specifically, in the analysis of carbon-electric energy storage power control strategy, the carbon emission cost of the charging and discharging power command set is first calculated using the following comprehensive cost calculation model:

[0084]

[0085] in: The total carbon emission cost at time t; Let be the dynamic coefficient of carbon price in the k-th time period; Energy exchange quantity; Carbon emission intensity factor; This is a power fluctuation penalty factor; This is the real-time power value; Energy conversion efficiency; For deep loop influence functions; The health status impact function is used. Data processing is performed based on this model, specifically: first, carbon emissions are calculated for the charging and discharging power commands at 96 time points, with the carbon emission coefficient of photovoltaic power generation set at 0.12 kg. / kWh, wind power set at 0.08kg / kWh. For a 5MWh energy storage device, with a photovoltaic power generation ratio of 70% and a wind power ratio of 30%, the carbon emissions for each time period are calculated. Threshold constraint analysis is then performed, setting the upper limit of carbon emission costs at 500 yuan per hour and the lower limit at 100 yuan per hour. When the carbon trading price is 45 yuan / ton, the cost constraints are obtained by calculating the carbon emission costs for each time period: peak-hour carbon emissions should not exceed 8 tons / hour, normal-hour emissions should not exceed 6 tons / hour, and valley-hour emissions should not exceed 4 tons / hour.

[0086] Based on these constraints, the charging and discharging power is adjusted. The adjustment is calculated using an iterative optimization method, with a calculation cycle of 15 minutes, gradually adjusting the charging and discharging power. When constraints are exceeded, the power is adjusted in increments of 0.2 MW until the constraints are met. A carbon-electricity linkage analysis is performed on the adjusted power sequence to examine the impact of power adjustments on carbon emissions, forming a carbon-electricity linkage feature set. During the scheduling strategy generation phase, the day is divided into 48 half-hour periods, and charging and discharging power is arranged for each period. For example, during peak photovoltaic power generation (10:00-15:00), the energy storage charging power is arranged at 75% of the rated power; during peak electricity consumption (18:00-21:00), the discharging power is arranged at 85% of the rated power. Technical constraint verification ensures that the strategy meets the operational requirements of the energy storage device.

[0087] Finally, the scheduling scheme was evaluated for performance, with evaluation indicators including: carbon emission reduction (tons / day), economic benefits (yuan / day), and equipment loss rate (% / day). The analytic hierarchy process (AHP) was used to determine the weights of each indicator: carbon emission reduction weight 0.4, economic benefits weight 0.35, and equipment loss weight 0.25. Taking a 4MWh energy storage device in an industrial park as an example, the specific implementation process during the 9:00-10:00 time period was as follows: the predicted output of photovoltaic power was 2.5MW, the predicted output of wind power was 1.2MW, and the load demand was 4.5MW. The initial charging power command was 1.8MW, and the calculated carbon emission was 1.2 tons / hour (0.8 tons from photovoltaic power and 0.4 tons from wind power). Considering a carbon price of 45 yuan / ton, the initial carbon emission cost was 54 yuan / hour. Since this did not exceed the cost constraint, no power correction was required. The generated scheduling strategy is as follows: 1.8MW charging power is executed from 9:00 to 9:30, and it is reduced to 1.5MW from 9:30 to 10:00. The SOC increases from 45% to 58%, achieving a carbon emission reduction of 0.9 tons and saving 40.5 yuan in costs.

[0088] In one specific embodiment, the process of performing step S106 may specifically include the following steps:

[0089] (1) Perform multi-objective decomposition on the carbon-electricity linkage energy storage scheduling strategy to obtain the target feature set, and perform target weight calculation on the target feature set to obtain the target weight vector;

[0090] (2) Perform multi-objective collaborative analysis on the target weight vector to obtain a collaborative optimization index set, and perform strategy adjustment based on the collaborative optimization index set to obtain an adjustment strategy sequence;

[0091] (3) Perform comprehensive performance evaluation on the adjustment strategy sequence to obtain strategy evaluation index, and perform scheme optimization based on the strategy evaluation index to obtain a set of candidate control strategies;

[0092] (4) The candidate control strategy set is evaluated by the collaborative evaluation algorithm to obtain the strategy evaluation results, and the carbon-electric energy storage power control strategy is constructed based on the strategy evaluation results.

[0093] Specifically, the energy storage dispatch strategy is decomposed into multiple objectives. The resulting objectives include indicators across three dimensions: economic objectives (minimizing operating costs and maximizing carbon trading revenue), environmental objectives (minimizing carbon emissions and maximizing renewable energy integration), and technical objectives (maximizing equipment lifespan and optimizing system stability). The weights of these objectives are calculated using the analytic hierarchy process (AHP), specifically by constructing a judgment matrix, calculating eigenvectors, and performing consistency checks. The initial weight allocation is: 0.35 for economic objectives, 0.4 for environmental objectives, and 0.25 for technical objectives. A multi-objective collaborative analysis is then performed on the objective weight vectors, employing a Pareto optimality-based multi-objective optimization method. First, a mapping relationship between objectives is established, and the coupling effect between different objectives is analyzed. For example, there is a certain conflict between economic and environmental objectives; when the carbon price is 45 yuan / ton, reducing carbon emissions by 1 ton requires an increase in operating costs of approximately 38 yuan. Based on these analysis results, a collaborative optimization index set is generated, including: comprehensive cost indicators, carbon emission intensity indicators, and equipment loss indicators.

[0094] The strategy is dynamically adjusted based on a collaborative optimization index set, using a rolling optimization method with a 15-minute optimization cycle. Within each cycle, the optimization objective is updated based on real-time data, and the charging and discharging strategy is adjusted accordingly. For example, during periods of high carbon prices (greater than 60 yuan / ton), the energy storage discharge power is appropriately increased to improve carbon emission reduction efficiency; when equipment wear is significant (cycle depth exceeds 20%), the charging and discharging power is reduced to extend equipment lifespan. A comprehensive performance evaluation is conducted on the adjusted strategy sequence, with evaluation indicators including daily average carbon emission reduction, economic benefits, and equipment lifespan loss rate. The evaluation process uses a fuzzy comprehensive evaluation method, combining quantitative and qualitative indicators for scoring. Based on the evaluation results, the top 5 schemes with the highest comprehensive scores are selected as the candidate control strategy set.

[0095] Finally, a collaborative evaluation algorithm is used to evaluate the candidate strategies, which comprehensively considers multiple evaluation indicators. A comprehensive score is calculated for each candidate strategy, taking into account the indicator weights and the degree of indicator achievement. The strategy with the highest comprehensive score is selected as the final carbon-electric energy storage power control strategy.

[0096] Taking a carbon-electric energy storage system in an industrial park as an example, the system includes 3MW of photovoltaic power generation, 2MW of wind power generation, and 4MWh of energy storage. In the multi-objective decomposition, the economic objectives include: operating costs controlled within 2000 yuan / day, and carbon trading revenue not less than 500 yuan / day; the environmental objectives include: reducing carbon emission intensity by 15%, and achieving a renewable energy consumption rate of 90%; the technical objectives include: energy storage lifespan loss rate not exceeding 0.2% / day, and power fluctuation rate controlled within 5%. Through multi-objective collaborative analysis, the specific implementation process during the 14:00-15:00 period is as follows: the actual photovoltaic output during this period is 2.8MW, the actual wind power output is 1.5MW, and the load demand is 5MW. The initial charging power is calculated to be 1.2MW. Considering the carbon price at that time was 55 yuan / ton, the economic benefit is 66 yuan / hour. After strategy adjustment, the charging power is reduced to 1.0MW. Although the economic benefit decreases slightly (by 8 yuan / hour), the equipment lifespan is better protected (cycle depth decreases from 22% to 18%). The final control strategy is as follows: execute 1.0MW charging power from 14:00 to 14:30, and maintain 0.8MW charging power from 14:30 to 15:00. Key performance indicators during strategy execution are: carbon emission reduction of 0.85 tons / hour, economic benefit of 58 yuan / hour, and equipment lifespan loss of 0.15% / day. This strategy achieves good carbon emission reduction while ensuring economic efficiency and the safe and stable operation of the equipment.

[0097] The above describes the analysis method for carbon-based electric energy storage power control strategy considering multi-source load synergy in the embodiments of this application. The following describes the analysis system for carbon-based electric energy storage power control strategy considering multi-source load synergy in the embodiments of this application. Please refer to [link to relevant documentation]. Figure 2 One embodiment of the carbon-electric energy storage power control strategy analysis system considering multi-source load coordination in this application includes:

[0098] The acquisition module 201 is used to perform multi-time-series feature extraction processing on the source-load power data of the carbon-electric energy storage system to obtain the source-load power collaborative feature matrix.

[0099] Processing module 202 is used to perform power balance analysis on the source-load power coordination feature matrix under carbon emission constraints to obtain the initial solution for power scheduling of the carbon-electric energy storage system.

[0100] Optimization module 203 is used to perform source-load collaborative optimization processing on the initial power scheduling solution to obtain a power optimization sequence that considers the complementary characteristics of multiple source loads;

[0101] The allocation module 204 is used to perform energy storage capacity allocation processing on the power optimization sequence to obtain the charging and discharging power instruction set of the energy storage unit;

[0102] The correction module 205 is used to perform power correction processing on the charging and discharging power instruction set under the constraint of carbon emission cost to obtain a carbon-electricity linked energy storage scheduling strategy.

[0103] The coordination module 206 is used to perform multi-objective coordinated optimization processing on the carbon-electric linkage energy storage scheduling strategy to obtain a carbon-electric energy storage power control strategy.

[0104] Through the collaborative efforts of the aforementioned components, and by performing multi-time-series feature extraction on the source-load power data of the carbon-electric energy storage system, the dynamic variation characteristics of source-load power can be comprehensively and accurately obtained, providing a reliable data foundation for subsequent collaborative optimization. Simultaneously, by performing power balance analysis under carbon emission constraints on the source-load power collaborative feature matrix, the system's power balance can be ensured while meeting carbon emission requirements, avoiding system instability caused by power imbalance. Furthermore, source-load collaborative optimization of the initial power scheduling solution fully considers the complementary characteristics between sources and loads, maximizing the advantages of various sources and loads and improving the overall operating efficiency of the system. In addition, by performing energy storage capacity allocation processing on the power optimization sequence… It can rationally allocate energy storage resources, avoid overuse or waste of energy storage capacity, and ensure the efficient utilization of energy storage units. Furthermore, by performing power correction processing on the charging and discharging power command set under carbon emission cost constraints, the charging and discharging strategy can be dynamically adjusted to adapt to changes in carbon emission costs, effectively reducing the operating cost of the system. Finally, by performing multi-objective collaborative optimization processing on the carbon-electric linkage energy storage scheduling strategy, a balanced optimization of multiple objectives such as economy, environmental protection, and reliability is achieved. This makes the final carbon-electric energy storage power control strategy highly practical and adaptable, which can not only ensure the stable operation of the system, but also maximize economic benefits while reducing carbon emissions, significantly improving the comprehensive performance and control effect of the carbon-electric energy storage system.

[0105] The above-described embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit them. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of this application.

Claims

1. A method for analyzing carbon-based electric energy storage power control strategies considering multi-source load synergy, characterized in that, The analytical method for carbon-based electric energy storage power control strategy considering multi-source load synergy includes: Multi-time-series feature extraction processing is performed on the source-load power data of the carbon-electric energy storage system to obtain the source-load power collaborative feature matrix. The power balance analysis under carbon emission constraints is performed on the source-load power coordination feature matrix to obtain the initial power scheduling solution of the carbon-electric energy storage system. This includes: mapping the source-load power coordination feature matrix to carbon emission factors to obtain a carbon emission feature dataset; performing time-series decomposition on the carbon emission feature dataset to obtain a carbon emission baseline sequence; extracting constraints from the carbon emission baseline sequence to obtain a carbon emission constraint boundary set; performing dynamic threshold calculation on the carbon emission constraint boundary set to obtain a carbon emission constraint threshold; constructing balance constraints between the carbon emission constraint threshold and power data to obtain power balance constraints; performing initial allocation of source-load power according to the power balance constraints to obtain a power allocation scheme; performing power balance verification on the power allocation scheme to obtain a feasible power scheduling solution set; and filtering the feasible power scheduling solution set to obtain the initial power scheduling solution of the carbon-electric energy storage system. The initial power scheduling solution is subjected to source-load cooperative optimization processing to obtain a power optimization sequence that considers the complementary characteristics of multiple source loads; The power optimization sequence is processed for energy storage capacity allocation to obtain the charging and discharging power instruction set of the energy storage unit; The charging and discharging power instruction set is subjected to power correction processing under carbon emission cost constraints to obtain a carbon-electricity linked energy storage scheduling strategy. The carbon-electricity linkage energy storage scheduling strategy is subjected to multi-objective collaborative optimization to obtain a carbon-electricity energy storage power control strategy.

2. The method for analyzing carbon-based electric energy storage power control strategies considering multi-source load synergy as described in claim 1, characterized in that, The process of extracting multiple time-series features from the source-load power data of the carbon-electric energy storage system yields a source-load power collaborative feature matrix, including: The historical power data of the carbon-electric energy storage system is divided into time windows to obtain a power data sequence of length T. The power data sequence is then standardized to obtain standardized power data, where T is a positive integer greater than 1. The standardized power data is decomposed by wavelet transform to obtain multi-scale time-frequency features of the power data, and the multi-scale time-frequency features are periodically analyzed by Fourier transform to obtain a power fluctuation feature sequence. The power fluctuation feature sequence is subjected to time-series correlation analysis to obtain the correlation feature vector of source load power, and the correlation feature vector is subjected to principal component analysis to obtain the dimensionality-reduced feature set. The power feature clustering process is performed on the dimensionality-reduced feature set by K-means clustering algorithm to obtain typical feature categories of source load power. Then, the power data is reconstructed based on the typical feature categories to obtain the reconstructed power sequence. The reconstructed power sequence is subjected to time-series feature encoding to obtain the time-series feature encoding matrix of source-load power, and a source-load power collaborative feature matrix is ​​constructed based on the time-series feature encoding matrix.

3. The method for analyzing carbon-based electric energy storage power control strategies considering multi-source load synergy as described in claim 1, characterized in that, The process of performing source-load cooperative optimization on the initial power scheduling solution to obtain a power optimization sequence considering the complementary characteristics of multiple source loads includes: The initial solution of power scheduling is subjected to source-load complementarity characteristic analysis to obtain the source-load complementarity relationship matrix. The source-load complementarity relationship matrix is ​​then subjected to time series segmentation to obtain time-segmented complementary feature sets. The source-load power matching degree is calculated on the time-segmented complementary feature set to obtain the source-load power matching degree index, and the power adjustment coefficient is generated based on the source-load power matching degree index to obtain the power adjustment coefficient sequence. The power regulation coefficient sequence is subjected to source-load collaborative optimization processing to obtain a collaborative optimization objective function, and the collaborative optimization objective function is weighted by an adaptive weight allocation algorithm to obtain optimized weight coefficients. The power scheduling scheme is adjusted collaboratively based on the optimized weight coefficients to obtain a source-load collaborative candidate sequence, and the source-load collaborative candidate sequence is evaluated for complementary performance to obtain a set of complementary performance indicators. The complementary performance index set is comprehensively evaluated to obtain the source-load synergistic optimization results, and a power optimization sequence considering the complementary characteristics of multiple source loads is constructed based on the source-load synergistic optimization results.

4. The method for analyzing carbon-based electric energy storage power control strategies considering multi-source load synergy as described in claim 1, characterized in that, The step of performing energy storage capacity allocation processing on the power optimization sequence to obtain the charging and discharging power instruction set of the energy storage unit includes: The energy storage capacity demand analysis is performed on the power optimization sequence to obtain the energy storage capacity demand feature set, and the energy storage capacity demand feature set is divided into time periods to obtain the time period capacity demand sequence. The energy storage unit status assessment process is performed on the time-segmented capacity demand sequence to obtain the energy storage status feature matrix, and the charge and discharge efficiency is calculated on the energy storage status feature matrix to obtain the charge and discharge efficiency coefficient. The energy storage capacity is initially allocated based on the charge and discharge efficiency coefficient to obtain an energy storage capacity allocation scheme. The energy storage capacity allocation scheme is then constrained and verified to obtain a feasible capacity allocation set. The feasible capacity allocation set is processed by charging and discharging power calculation to obtain a charging and discharging power sequence, and power command generation is performed based on the charging and discharging power sequence to obtain the charging and discharging power command set of the energy storage unit.

5. The method for analyzing carbon-based electric energy storage power control strategies considering multi-source load synergy as described in claim 1, characterized in that, The process of performing power correction processing on the charging and discharging power instruction set under carbon emission cost constraints to obtain a carbon-electricity linked energy storage scheduling strategy includes: Carbon emission cost calculation is performed on the charge and discharge power instruction set to obtain a carbon emission cost sequence, and threshold constraint analysis is performed on the carbon emission cost sequence to obtain cost constraint conditions. The charging and discharging power is corrected according to the cost constraints to obtain a power correction sequence, and the power correction sequence is subjected to carbon-electric linkage analysis to obtain a carbon-electric linkage feature set. The carbon-electric linkage feature set is processed to generate a scheduling strategy, resulting in a carbon-electric linkage scheduling scheme. The feasibility of the carbon-electric linkage scheduling scheme is then verified to obtain a set of feasible scheduling schemes. The feasible scheduling scheme set is subjected to scheduling performance evaluation to obtain scheduling performance indicators, and the scheduling schemes are screened according to the scheduling performance indicators to obtain a carbon-electricity linkage energy storage scheduling strategy.

6. The method for analyzing carbon-based electric energy storage power control strategies considering multi-source load synergy as described in claim 1, characterized in that, The multi-objective collaborative optimization process of the carbon-electric linkage energy storage scheduling strategy yields a carbon-electric energy storage power control strategy, including: The carbon-electricity linkage energy storage scheduling strategy is decomposed into multiple objectives to obtain a set of objective features, and the objective feature set is then weighted to obtain an objective weight vector. The target weight vector is subjected to multi-objective collaborative analysis to obtain a collaborative optimization index set, and a strategy adjustment process is performed based on the collaborative optimization index set to obtain an adjustment strategy sequence. The adjustment strategy sequence is subjected to comprehensive performance evaluation to obtain strategy evaluation indexes, and a scheme optimization process is performed based on the strategy evaluation indexes to obtain a set of candidate control strategies. The candidate control strategy set is evaluated using a collaborative evaluation algorithm to obtain the strategy evaluation results, and a carbon-electric energy storage power control strategy is constructed based on the strategy evaluation results.

7. A carbon-based electric energy storage power control strategy analysis system considering multi-source load synergy, used to implement the carbon-based electric energy storage power control strategy analysis method considering multi-source load synergy as described in any one of claims 1-6, characterized in that, The carbon-based electric energy storage power control strategy analysis system considering multi-source load coordination includes: The acquisition module is used to perform multi-time-series feature extraction processing on the source-load power data of the carbon-electric energy storage system to obtain the source-load power collaborative feature matrix. The processing module is used to perform power balance analysis under carbon emission constraints on the source-load power coordination feature matrix to obtain an initial solution for power scheduling of the carbon-electric energy storage system. This includes: mapping the source-load power coordination feature matrix to carbon emission factors to obtain a carbon emission feature dataset; performing time-series decomposition on the carbon emission feature dataset to obtain a carbon emission baseline sequence; extracting constraints from the carbon emission baseline sequence to obtain a set of carbon emission constraint boundaries; performing dynamic threshold calculation on the set of carbon emission constraint boundaries to obtain a carbon emission constraint threshold; constructing balance constraints between the carbon emission constraint threshold and power data to obtain power balance constraints; performing initial allocation of source-load power according to the power balance constraints to obtain a power allocation scheme; performing power balance verification on the power allocation scheme to obtain a set of feasible power scheduling solutions; and filtering the set of feasible power scheduling solutions to obtain the initial solution for power scheduling of the carbon-electric energy storage system. The optimization module is used to perform source-load collaborative optimization processing on the initial power scheduling solution to obtain a power optimization sequence that considers the complementary characteristics of multiple source loads. The allocation module is used to perform energy storage capacity allocation processing on the power optimization sequence to obtain the charging and discharging power instruction set of the energy storage unit; The correction module is used to perform power correction processing on the charging and discharging power instruction set under the constraint of carbon emission cost, so as to obtain a carbon-electricity linked energy storage scheduling strategy. The coordination module is used to perform multi-objective collaborative optimization processing on the carbon-electric linkage energy storage scheduling strategy to obtain the carbon-electric energy storage power control strategy.

Images