A user-side green electricity optimization scheduling method and system

Abstract

The application discloses a user-side green electricity optimization scheduling method and system, applied to the technical field of energy optimization scheduling, comprising the following steps: obtaining active power data of a power distribution network and power load data of each power consumption device; based on the first time sequence characteristics of the active power data, obtaining a source-side characteristic vector sequence, and based on the second time sequence characteristics of each power load data, obtaining a load-side characteristic vector sequence; determining source-load curve similarity information; based on the source-load curve similarity information and schedulable power information, obtaining a green electricity matching index; constructing a green electricity optimization scheduling model, determining the constraint conditions of the green electricity optimization scheduling model, solving the green electricity optimization scheduling model according to the constraint conditions and the green electricity matching index, and obtaining a green electricity optimization scheduling scheme; and scheduling the green electricity power of each power consumption device on the user side according to the green electricity optimization scheduling scheme. The method can improve the green electricity consumption rate and guarantee the safe and stable operation of the power distribution network.

Description

Technical Field

[0001] This invention relates to the field of energy optimization and dispatching technology, and in particular to a user-side green electricity optimization and dispatching method and system. Background Technology

[0002] Renewable energy power generation is intermittent and highly volatile. By rationally arranging the power consumption of user-side electrical equipment, it is crucial to improve the green electricity consumption rate and ensure the safe and stable operation of the distribution network.

[0003] In existing technologies, green electricity dispatching schemes are developed by predicting renewable energy generation and then matching the predictions with historical electricity loads on the user side. However, due to the strong randomness of renewable energy generation and the significant uncertainty of user-side electricity loads, the green electricity consumption rate of the dispatching schemes obtained using these methods is low in practical applications. Furthermore, when actual renewable energy generation or electricity load deviates from the predicted values, the pre-defined dispatching scheme loses its optimality and may even become infeasible, making it difficult to guarantee the robustness and real-time executability of the green electricity dispatching scheme, thus affecting the safe and stable operation of the distribution network. Summary of the Invention

[0004] This invention provides a user-side green electricity optimization scheduling method and system to solve the technical problem of low green electricity absorption rate in existing green electricity scheduling methods, so as to improve the green electricity absorption rate and ensure the safe and stable operation of the distribution network.

[0005] To address the aforementioned technical problems, in a first aspect, the present invention provides a user-side green electricity optimized scheduling method, comprising: Obtain active power data of the power distribution network and power load data of each electrical device; Based on the first time-series characteristics of the active power data, a source-side feature vector sequence is obtained, and based on the second time-series characteristics of each of the electricity load data, a load-side feature vector sequence is obtained. Based on the source-side feature vector sequence and the load-side feature vector sequence, the source-load curve similarity information is determined; Based on the source-load curve similarity information and the dispatchable power information, a green electricity matching index is obtained, wherein the dispatchable power information is determined by the operating baseline data of the distribution network; Using the highest green electricity absorption rate as the objective function, a green electricity optimization scheduling model is constructed. Based on green electricity supply constraints, dispatchable load constraints, total user electricity demand constraints, and distribution network security constraints, the constraints of the green electricity optimization scheduling model are determined. Based on the constraints and the green electricity matching index, the green electricity optimization scheduling model is solved to obtain the green electricity optimization scheduling scheme. According to the green electricity optimization scheduling scheme, the green electricity power of each of the user-side electrical devices is scheduled.

[0006] As one preferred embodiment, the step of obtaining the source-side feature vector sequence based on the first time-series features of the active power data includes: The active power data is standardized to obtain standard active power data; Based on the aforementioned standard active power data, the source-side power curve is determined; The source-side power curve is processed to obtain the first timing feature; The first time-series feature is analyzed to obtain the source-side feature vector sequence.

[0007] As one preferred embodiment, the step of obtaining the load-side feature vector sequence based on the second time-series characteristics of each of the electricity load data includes: The power load data of each of the aforementioned electrical devices is standardized to obtain standard power load data. Based on all the aforementioned standard electricity load data, determine the load-side electricity load curve; The load curve on the load side is processed to obtain the second time-series characteristic; The second time-series feature is analyzed to obtain the load-side feature vector sequence.

[0008] As one preferred embodiment, determining the source-load curve similarity information based on the source-side feature vector sequence and the load-side feature vector sequence includes: A first similarity analysis is performed on the source-side feature vector sequence and the load-side feature vector sequence to obtain first morphological similarity information; A second similarity analysis is performed on the source-side feature vector sequence and the load-side feature vector sequence to obtain second feature similarity information; A third similarity analysis is performed on the source-side power curve and the load-side electrical load curve to obtain third amplitude similarity information; The source-load curve similarity information is obtained by integrating the first morphological similarity information, the second feature similarity information, and the third amplitude similarity information.

[0009] As one preferred embodiment, the step of performing a first similarity analysis on the source-side feature vector sequence and the load-side feature vector sequence to obtain first morphological similarity information includes: Based on the Euclidean distance calculation principle, each feature vector in the source-side feature vector sequence and each feature vector in the corresponding load-side feature vector sequence are processed to obtain the source-load feature distance matrix; The source-load feature distance matrix is ​​processed based on the dynamic time warping algorithm, and the processing result is analyzed to obtain the first morphological similarity information.

[0010] As one preferred embodiment, the step of performing a second similarity analysis on the source-side feature vector sequence and the load-side feature vector sequence to obtain second feature similarity information includes: Based on the source-side feature vector sequence, a high-dimensional representation of the source-side feature vector is obtained, and based on the load-side feature vector sequence, a high-dimensional representation of the load-side feature vector is obtained. Based on the principle of cosine similarity, the high-dimensional representation results of the source-side feature vector and the high-dimensional representation results of the load-side feature vector are processed to obtain the second feature similarity information.

[0011] As one preferred embodiment, the third similarity analysis of the source-side power curve and the load-side electricity consumption curve to obtain third amplitude similarity information includes: The source-side power curve and the load-side electricity consumption curve are time-aligned respectively to obtain the corresponding source-side power curve and load-side electricity consumption curve to be analyzed. Based on the pre-constructed amplitude matching degree model, the power curve on the source side and the electricity load curve on the load side to be analyzed are processed to obtain the third amplitude similarity information.

[0012] As one preferred embodiment, the process of obtaining the green electricity matching index based on the source-load curve similarity information and schedulable power information includes: The load-side electrical load curve is analyzed to obtain load adjustability margin information; Based on the operating baseline data and the load adjustability margin information, the dispatchable power information is determined; The source-load curve similarity information and the schedulable power information are integrated to obtain the green electricity matching index.

[0013] As one preferred embodiment, the step of scheduling the green power of each of the user-side electrical devices according to the green power optimization scheduling scheme includes: Based on the green electricity optimization scheduling scheme, green electricity power scheduling instructions corresponding to each of the electrical devices are generated. According to the green power scheduling instruction, the green power of each of the electrical devices on the user side is scheduled.

[0014] To address the aforementioned technical problems, this invention also provides a user-side green electricity optimization scheduling system, comprising: The data acquisition module is used to acquire active power data of the power distribution network and power load data of various electrical devices; The feature vector module is used to obtain a source-side feature vector sequence based on the first time-series features of the active power data, and to obtain a load-side feature vector sequence based on the second time-series features of each of the electrical load data. The similarity analysis module is used to determine the source-load curve similarity information based on the source-side feature vector sequence and the load-side feature vector sequence. The green electricity matching index module is used to obtain the green electricity matching index based on the source-load curve similarity information and the dispatchable power information, wherein the dispatchable power information is determined by the operating reference data of the distribution network; The scheduling scheme module is used to construct a green electricity optimization scheduling model with the highest green electricity consumption rate as the objective function, and to determine the constraints of the green electricity optimization scheduling model based on green electricity supply constraints, dispatchable load constraints, total user electricity demand constraints, and distribution network security constraints. Based on the constraints and the green electricity matching index, the green electricity optimization scheduling model is solved to obtain the green electricity optimization scheduling scheme. The scheduling execution module is used to schedule the green power of each of the power-consuming devices on the user side according to the green power optimization scheduling scheme.

[0015] Compared with existing technologies, the beneficial effects of the embodiments of the present invention are at least one of the following: The present invention achieves deep extraction of multi-dimensional morphological features of the source-load curve by performing time-series feature analysis on the acquired active power data of the distribution network to obtain the source-side feature vector sequence, and performing time-series feature analysis on the load data of each electrical device to obtain the load-side feature vector sequence; The present invention achieves quantitative comparison of the green power output curve and the user-side load curve in terms of trend, fluctuation mode and local features by determining the similarity information of the source-load curve, thereby solving the technical defects of traditional Euclidean distance or simple correlation coefficient that cannot tolerate local scaling of the time axis and are difficult to accurately measure the consistency of source-load morphology; The present invention constructs a green power matching index based on the similarity information of the source-load curve and the dispatchable power information determined by the distribution network operation benchmark data, thereby integrating "source-load morphology matching potential" and "actual dispatchable capacity of the power grid" into a single comprehensive decision index, thereby solving the one-sidedness of dispatch decision caused by existing methods either ignoring the power grid carrying capacity or failing to quantify the degree of matching. The present invention addresses the following issues: It uses the highest green energy absorption rate as the objective function and simultaneously considers green energy supply constraints, dispatchable load constraints, total user electricity demand constraints, and distribution network security constraints to construct an optimized scheduling model. This achieves mathematical modeling for optimizing green energy utilization under multiple physical and user boundary conditions. Furthermore, by solving the green energy optimization scheduling model based on constraints and a green energy matching index, the present invention obtains an optimized green energy scheduling scheme. This strategy guides the priority allocation of green energy power during periods of high matching degree and high margin, thus solving the problems of low green energy absorption rate and wasted high green energy periods caused by the lack of spatiotemporal priority quantification indicators in existing technologies. Finally, the present invention schedules the green energy power of various user-side electrical equipment according to the optimized green energy scheduling scheme, achieving closed-loop linkage between source-load matching analysis and actual control commands. This solves the lag problem of traditional scheduling schemes, which only remain at the planning level and cannot be dynamically adjusted based on real-time operating baseline data. This improves the robustness and real-time executability of the green energy scheduling scheme, thereby contributing to ensuring the safe and stable operation of the distribution network. Attached Figure Description

[0016] Figure 1 This is a flowchart illustrating a user-side green electricity optimization scheduling method in one embodiment of the present invention; Figure 2 This is a schematic diagram of the structure of a user-side green electricity optimization scheduling system in one embodiment of the present invention; Figure label: The module includes: 11. Data acquisition module; 12. Feature vector module; 13. Similarity analysis module; 14. Green electricity matching index module; 15. Scheduling scheme module; and 16. Scheduling execution module. Detailed Implementation

[0017] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. The purpose of providing these embodiments is to make the disclosure of the present invention more thorough and comprehensive. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of the present invention.

[0018] In the description of this application, the terms "first," "second," "third," etc., are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of indicated technical features. Therefore, a feature defined with "first," "second," "third," etc., may explicitly or implicitly include one or more of that feature. In the description of this application, unless otherwise stated, "a plurality of" means two or more.

[0019] In the description of this application, it should be noted that, unless otherwise expressly specified and limited, the terms "installation," "connection," and "linking" should be interpreted broadly. For example, they can refer to fixed connections, detachable connections, or integral connections; they can refer to mechanical connections or electrical connections; they can refer to direct connections or indirect connections through an intermediate medium; and they can refer to the internal communication between two components. The terms "vertical," "horizontal," "left," "right," "upper," "lower," and similar expressions used herein are for illustrative purposes only and do not indicate or imply that the device or component referred to must have a specific orientation, or be constructed and operated in a specific orientation, and therefore should not be construed as limiting the invention. The term "and / or" as used herein includes any and all combinations of one or more of the associated listed items. Those skilled in the art can understand the specific meaning of the above terms in this application based on the specific circumstances.

[0020] In the description of this application, it should be noted that, unless otherwise defined, all technical and scientific terms used in this invention have the same meaning as commonly understood by one of ordinary skill in the art. The terminology used in this specification is for the purpose of describing specific embodiments only and is not intended to limit the invention. Those skilled in the art can understand the specific meaning of the above terms in this application based on the specific circumstances.

[0021] Against the backdrop of continuously increasing demands for large-scale distributed energy integration and efficient distribution network operation, renewable energy generation, represented by photovoltaic and wind power, has become an important source of power supply for distribution networks. However, renewable energy power generation exhibits significant intermittency, random fluctuations, and non-stationarity, with its output fluctuating in real time with natural conditions, making it difficult to maintain stability and control. Simultaneously, user-side electrical equipment is diverse and operates under complex conditions, with electricity loads dynamically changing with production and daily life activities, also exhibiting strong uncertainty and time-varying characteristics. Under these conditions, scientifically and rationally scheduling strategies to coordinate the power consumption of user-side electrical equipment, achieving efficient synergy between renewable energy generation and user-side electricity load, plays a crucial role in improving green electricity absorption, maintaining power balance in the distribution network, and ensuring the safe and stable operation of the distribution network.

[0022] Existing user-side green electricity dispatching technologies generally adopt the following implementation method: First, the power generation of renewable energy is predicted to obtain the power generation forecast results for future periods; second, historical electricity load data of the user side is retrieved, and the predicted power generation results of renewable energy are matched and compared with the historical electricity load; finally, a green electricity dispatching scheme is formulated based on the above matching relationship, and green electricity power dispatching is performed on user-side electrical equipment according to the established scheme.

[0023] However, this dispatching method has obvious technical shortcomings: On the one hand, renewable energy output is highly random, and user-side electricity load has significant uncertainty. Relying solely on power generation forecasts and historical load matching cannot accurately depict the dynamic changes on both the source and load sides. The matching method is crude and lacks accuracy, resulting in a low green electricity consumption rate in actual operation of the dispatching scheme, making it difficult to fully utilize renewable energy power generation resources. On the other hand, the traditional scheme is a pre-formulated static dispatching plan, which lacks real-time adaptive adjustment capabilities. When the actual renewable energy power generation or real-time electricity load deviates from the forecast value, the original dispatching scheme quickly loses its optimality, and may even encounter situations such as constraint overruns and scheme infeasibility. It cannot guarantee the robustness and real-time executability of the dispatching scheme, which can easily cause problems such as power imbalance and voltage anomalies in the distribution network, directly affecting the safe and stable operation of the distribution network.

[0024] To address the aforementioned technical problems, one embodiment of the present invention provides a user-side green energy optimization scheduling method. For details, please refer to [link to relevant documentation]. Figure 1 , Figure 1 The diagram shown is a flowchart of a user-side green electricity optimization scheduling method according to one embodiment of the present invention. The method includes steps S1 to S6: S1. Obtain active power data of the power distribution network and power load data of each electrical device; S2. Based on the first time-series characteristics of the active power data, a source-side feature vector sequence is obtained, and based on the second time-series characteristics of each of the electrical load data, a load-side feature vector sequence is obtained. S3. Based on the source-side feature vector sequence and the load-side feature vector sequence, determine the source-load curve similarity information; S4. Based on the source-load curve similarity information and the dispatchable power information, a green electricity matching index is obtained, wherein the dispatchable power information is determined by the operating reference data of the distribution network; S5. Using the highest green electricity consumption rate as the objective function, construct a green electricity optimization scheduling model. Based on green electricity supply constraints, dispatchable load constraints, total user electricity demand constraints, and distribution network security constraints, determine the constraints of the green electricity optimization scheduling model. Solve the green electricity optimization scheduling model based on the constraints and the green electricity matching index to obtain the green electricity optimization scheduling scheme. S6. According to the green electricity optimization scheduling scheme, the green electricity power of each of the power-consuming devices on the user side is scheduled.

[0025] In step S1, both the green electricity output on the source side and the electricity load on the load side exhibit strong randomness, time-varying nature, and nonlinearity. The temporal fluctuation patterns, peak and valley occurrence times, and power variation amplitudes of these two factors directly determine the accuracy of green electricity matching and the feasibility of scheduling. Only by synchronously and in real-time acquiring the active power data on the distribution network side and the electricity load data of each user's electrical equipment can we provide real, complete, and synchronous basic data support for subsequent time-series feature extraction, source-load curve similarity calculation, green electricity matching index construction, and optimization scheduling model solution. This avoids the scheduling scheme from becoming distorted or unexecutable due to missing, delayed, or incomplete data.

[0026] Specifically, firstly, the overall active power of the distribution network is acquired using the metering points at the distribution gateway and the grid-connected points of distributed new energy sources as data acquisition nodes; the power load of individual electrical devices is acquired using the intelligent monitoring terminals, smart meters, and control cabinets of each user-side electrical equipment as data acquisition nodes. Secondly, a periodic real-time acquisition method is adopted, with synchronous sampling at fixed time intervals (e.g., 5 minutes, 15 minutes); the data is uploaded to the dispatch core unit through the power Internet of Things, RS485, Ethernet, power line carrier, or wireless communication modules to ensure strict alignment of source and load data timestamps. Then, the acquired raw data is denoised, imputed, time-stamped aligned, and outlier removed to form a standardized time-series dataset, ensuring that the data quality meets the requirements for feature extraction and model calculation. Finally, the processed active power time-series data and the load time-series data of each electrical device are stored in their respective data queues as direct input data for subsequent feature vector modules and similarity analysis modules.

[0027] The electrical equipment in this invention includes user-side rigid load equipment, flexible adjustable load equipment, and aggregated load entities. The aggregated load entity is a group of similar equipment entities participating in the scheduling, such as an electric vehicle aggregator, used to centrally coordinate and optimize the charging power of electric vehicles within its jurisdiction.

[0028] In step S2, both the active power of renewable energy generation and the electricity load on the user side are typical non-stationary time-series data. Using only the raw power values ​​cannot characterize their inherent operating patterns such as fluctuation trends, peak-valley phases, rates of change, and local abrupt changes, making it difficult to achieve accurate source-load curve matching. To deeply explore the time-series patterns, multi-scale fluctuations, and dynamic correlation characteristics of source-side output and load-side load, and to overcome the low accuracy and poor robustness caused by traditional matching based solely on power amplitude, it is necessary to extract the first time-series features of the active power data and the second time-series features of the electricity load data, respectively, and transform them into a standardized, high-dimensional feature vector sequence. This provides a unified and quantifiable feature foundation for subsequent source-load curve similarity calculations and the construction of the green electricity matching index.

[0029] Specifically, firstly, the active power data of the distribution network and the load data of each electrical device are subjected to maximum-minimum normalization to eliminate differences in dimensions and numerical magnitudes, mapping the data to a unified interval to obtain standard active power data and standard load data, avoiding the obscuring of time-series characteristics by amplitude differences. Secondly, based on the standardized time-series data, continuously changing source-side power curves and load-side load curves are constructed with time as the horizontal axis and power or load value as the vertical axis, forming a complete time-series curve that intuitively reflects the changes in power or load over time. Next, wavelet transform algorithm is used to perform multi-scale analysis on the source-side power curve to extract operating characteristics such as trend term, fluctuation term, peak value, valley value, ramp rate, and phase, forming the first time-series feature; similarly, wavelet transform algorithm is used to synchronously analyze the load-side load curve to extract operating characteristics such as overall load trend, adjustability margin, fluctuation frequency, peak and valley periods, and demand rigidity, forming the second time-series feature. Finally, the first temporal feature corresponding to each time step is vectorized and encoded, and arranged in chronological order to form a source-side feature vector sequence. In addition, the second temporal feature corresponding to each time step is vectorized and encoded, and arranged in chronological order to form a load-side feature vector sequence.

[0030] The source-side power curve is a continuous time-series curve with time as the independent variable and the active power of distributed renewable energy (photovoltaics and wind power) in the distribution network as the dependent variable. It fully reflects the magnitude, fluctuation pattern, and peak-valley distribution of green electricity output within the dispatch cycle, and is a direct expression of energy supply on the source side. The load-side electricity consumption curve is a continuous time-series curve with time as the independent variable and the total electricity consumption of all electrical equipment on the user side (including electric vehicle charging load, air conditioning, heat pumps, energy storage, industrial equipment, etc.) as the dependent variable. It reflects the changing pattern of total electricity demand on the user side over time and includes both rigid loads and flexible adjustable loads.

[0031] Preferably, this invention provides a source-side power curve generation method for the distribution network corresponding to electric vehicle aggregators. First, a sequence of net active power prediction values ​​with time intervals of Δt, provided by a meteorological station or a local renewable energy prediction model, is obtained, and this sequence is then organized into a source-side power prediction curve. Where t = 1, 2, ..., N. Additionally, the real-time active power of renewable energy is obtained from the real-time monitoring system of the renewable energy power plant. Using real-time active power as the benchmark for rolling optimization and deviation correction, and in order to eliminate the impact of prediction deviation on subsequent curve matching, prediction error calculation is performed at each rolling time step: This error term reflects the deviation of the current net active power prediction result from the actual active power data. To control the decay rate of the error correction, a time decay correction coefficient is introduced: in, When t=t k hour, This indicates the largest correction magnitude.

[0032] The source-side power prediction curve is rolled over to obtain the corrected source-side power curve: In step S3, the source-side feature vector sequence and the load-side feature vector sequence only represent the time-series characteristics of green electricity output and electricity load, respectively. They cannot directly reflect the degree of matching between the two in terms of morphological trends, characteristic correlations, and power amplitudes, making it difficult to support subsequent green electricity matching index construction and optimized scheduling decisions. To achieve accurate quantitative comparison of the source and load sides in the time-series dimension and overcome the shortcomings of traditional single numerical comparison in handling local curve scaling, phase shifts, and fluctuation misalignments, it is necessary to conduct multi-dimensional fusion analysis based on the source-side and load-side feature vector sequences to obtain source-load curve similarity information that can comprehensively reflect the degree of matching between the source and load curves, providing a core decision-making basis for optimized green electricity scheduling.

[0033] Specifically, firstly, a distance matrix of the source-load feature vector sequences is constructed based on Euclidean distance. Then, the optimal warping path is searched using the Dynamic Time Warping (DTW) algorithm, and the minimum cumulative distance is calculated. This distance is then mapped to a similarity value within the [0,1] interval to obtain the first morphological similarity information, which characterizes the similarity of the source-load curves in terms of overall trend and fluctuation shape. Secondly, the source-side and load-side feature vector sequences are represented in high-dimensional space. The cosine value of the angle between the high-dimensional vectors is calculated using the cosine similarity principle to obtain the second feature similarity information, which characterizes the directional consistency of the source and load in terms of multi-scale wavelet features, trend components, and fluctuation patterns. Next, the source-side power curve and the load-side electricity consumption curve are time-series aligned. Then, the matching ratio between the power amplitude and the load amplitude at the same moment is calculated using the amplitude matching degree model to obtain the third amplitude similarity information, which characterizes the matching degree of the source and load in terms of power scale and electricity supply capacity. Finally, the first morphological similarity information, the second feature similarity information, and the third amplitude similarity information are weighted and fused to obtain a comprehensive and unique source-load curve similarity information, which serves as the core input for subsequent green electricity matching index calculation.

[0034] Among them, the first morphological similarity information refers to the similarity index calculated based on the Dynamic Time Warping (DTW) algorithm and Euclidean distance, which is used to quantify the overall shape, fluctuation trend, and peak-valley phase consistency of the source-load curve. It focuses on measuring whether the curves "look similar" and can tolerate local scaling and translation of the time axis, and is not affected by the absolute magnitude of the amplitude.

[0035] Preferably, firstly, the source-side feature vector sequence is calculated using the Euclidean distance calculation principle. and load side feature vector sequence The local distance between the feature vectors of any two time points is used to obtain an N×N source-load feature distance matrix. The principle of Euclidean distance calculation is as follows: Secondly, the DTW algorithm is used to find an optimal regularized path from (1,1) to (N,N) through dynamic programming. , where represents the i-th feature vector sequence of the source side. k Point and load-side eigenvector sequence j-th k Point alignment. Minimum cumulative distance corresponding to the optimal path. This refers to the morphological difference between two curves. To obtain a similarity index within the range [0,1], the following mapping is performed to generate a temporal morphological similarity sequence. : in, The local distances corresponding to all load-side points aligned with the source-side time t in the optimal regularized path W. The average value; is the standard deviation of all local distances. The closer it is to 1, the more similar the curve shape is to time t.

[0036] The second feature similarity information refers to the similarity index calculated based on high-dimensional feature vectors and cosine similarity. It is used to quantify the inherent consistency of source-load curves in multi-scale wavelet features, local abrupt changes, and trend components, and is insensitive to noise and amplitude changes.

[0037] Preferably, the cosine similarity principle is shown in the following formula: By calculating the cosine similarity between the source-side feature vector sequence and the load-side feature vector sequence, the second feature similarity information between the two is obtained.

[0038] The third amplitude similarity information refers to the similarity index used to quantify the matching degree between source-side power and load-side load in terms of real-time power scale and power supply capacity after time-series alignment, reflecting the actual support capacity of green power output for electricity load.

[0039] Preferably, firstly, the source-side power curve and the load-side electrical load curve are time-series aligned to obtain the source-side power curve and the load-side electrical load curve to be analyzed. Secondly, an amplitude matching degree model is constructed based on the matching degree calculation formula, which is: in, To analyze the source-side power curve, To analyze the load curve on the load side, A very small positive constant (e.g., 1e-6) is used to prevent division by zero errors. Finally, the power curve on the source side and the load curve on the load side to be analyzed are input into the amplitude matching model to obtain the third amplitude similarity information.

[0040] In step S4, the source-load curve similarity information can only reflect the matching potential between the source-side output and the load-side load in terms of temporal patterns, and cannot reflect the actual power regulation capability and safe operation boundary of the distribution network. If scheduling is based solely on similarity, it is easy to encounter situations where the matching degree is high but the grid has no adjustment space and the load has no adjustment margin, leading to problems such as the scheduling scheme being unexecutable and exceeding the limits. Therefore, it is necessary to integrate the source-load matching potential with the actual dispatchable capability of the grid to construct a comprehensive decision index that combines "matching degree" and "executability," namely the Green Electricity Matching Index (GLMI), providing a single quantitative basis that can be directly used for calculation in subsequent optimized scheduling models.

[0041] Specifically, firstly, based on the load-side electricity demand curve, rigid electricity demand and flexible adjustable space are distinguished to calculate the load adjustability margin information. Secondly, combined with the distribution network operation benchmark data (including benchmark load, transformer capacity, line limits, voltage constraints, etc.), and superimposed with the load adjustability margin information, the upper limit and range of power that the system can dispatch within a safe range are determined to form dispatchable power information. Finally, the source-load curve similarity information and dispatchable power information are weighted, fused, and normalized to obtain a comprehensive quantitative index within the range of [0,1], namely the green electricity matching index; the higher the index, the more suitable it is to prioritize the consumption of green electricity during that period.

[0042] Among them, load adjustability margin information refers to the range and time interval within which user-side electrical equipment can increase, decrease, or shift power consumption upwards or downwards while meeting basic electricity needs. It is a core parameter characterizing load adjustment potential. Dispatchable power information refers to the range of green electricity power that the dispatching system can actually adjust and allocate under the joint constraints of distribution network safety operation constraints and user electricity consumption constraints. It is determined by the distribution network operation benchmark data (benchmark load, capacity limit, safety margin) and load adjustability margin information, and is the physical boundary by which dispatching instructions can be implemented. The green electricity matching index is a comprehensive decision-making quantitative indicator that integrates source-load curve similarity and dispatchable power. It is used to intuitively and uniformly evaluate the suitability of implementing green electricity priority dispatch in a certain period. The higher the index, the better the source-load form is matched and the more sufficient the grid adjustment space is, and the more priority should be given to allocating green electricity. The lower the index, the poorer the matching or the insufficient adjustability, and the less green electricity should be allocated.

[0043] Preferably, the present invention provides a method for calculating dispatchable power information. First, the distribution network safety margin is calculated, which ensures that the dispatched electric vehicle charging power does not violate the physical limitations of the grid side. The distribution network safety margin is calculated using the following formula: in, This refers to the capacity of the main transformer in the distribution network. For critical path current carrying limit, As the baseline load, This is a safety buffer value.

[0044] Secondly, calculate the EV load flexibility margin, which is the total amount of EV load that the electric vehicle aggregator can dispatch or reduce at time t. The EV load flexibility margin is calculated using the following formula: in, As the attenuation factor, This is the load curve for the load-side users.

[0045] Finally, the minimum values ​​of the distribution network safety margin and EV load flexibility margin at time t are used as the dispatchable power information.

[0046] And only when Only when the value is greater than 0 can the physical basis for executing scheduling and absorbing green electricity be established.

[0047] Preferably, the green electricity allocation index at time t is calculated by the following formula: in, For the calculated source-load curve similarity information; It is the calculated schedulable power information. The normalized value (mapped to the [0,1] interval). α and β are the technology matching weight and execution capability weight, and α+β=1. It is the green electricity priority supply intensity factor.

[0048] In step S5, the source-load curve similarity and green electricity matching index can only characterize the suitability and priority of green electricity consumption in each time period, and cannot directly form an executable scheduling instruction. To maximize green electricity utilization efficiency while meeting the physical constraints of the power grid, user electricity demand, and equipment operating limitations, the highest green electricity consumption rate must be taken as the optimization objective. A mathematical programming model is constructed by combining multiple boundary constraints, and the green electricity matching index is incorporated as a scheduling priority guiding factor into the model solution. This generates a green electricity optimized scheduling scheme that considers consumption effect, execution feasibility, and power grid security, achieving a closed loop from analysis and decision-making to control execution.

[0049] Specifically, firstly, the objective function is to maximize the green electricity consumption rate over the entire scheduling cycle. The green electricity consumption, total electricity consumption, and green electricity matching index are weighted and incorporated into the objective expression to guide the model to prioritize scheduling strategies with high matching and high consumption efficiency. Secondly, constraints are determined: green electricity supply constraints ( This refers to the dispatching of green electricity not exceeding the current actual supply capacity of renewable energy; dispatchable load constraints ( This refers to the fact that the power dispatched by each electrical device does not exceed its adjustable upper and lower limits, and does not violate the flexible load adjustment rules; the total power demand constraint of users ( This refers to ensuring that the total electricity demand of users is fully met within the dispatching cycle; distribution network security constraints ( This refers to meeting the safety constraints of distribution network operation, such as transformer capacity, line current carrying capacity, voltage deviation, and power flow balance. Finally, the green electricity matching index is substituted into the green electricity optimization scheduling model as a guiding factor, and a mixed integer programming algorithm is used to solve it, obtaining the green electricity optimization scheduling scheme for each time period and each device.

[0050] The green electricity optimization scheduling model is a mathematical optimization model with multiple constraints and a single objective. The overall architecture includes: the objective layer (with the highest green electricity consumption rate as the core objective), the decision variable layer (the green electricity usage power and scheduling power of each power-consuming device in each time period), the constraint layer (green electricity supply constraints, dispatchable load constraints, total electricity demand constraints of users, and distribution network security constraints), and the guidance layer (green electricity matching index, used to improve the scheduling priority during high matching periods).

[0051] Preferably, the present invention employs a hierarchical master-slave objective structure to construct a green energy optimization scheduling model, with the master objective being to minimize the difference between the planned green energy trading curve x(t) and the predicted green energy output curve. The dynamic time warping (DTW) distance between them.

[0052] The goal is to maximize the total GLMI-weighted green energy consumption over the entire scheduling cycle.

[0053] In step S6, in order to achieve closed-loop execution of source-load matching analysis, model optimization calculation and on-site power regulation, and to ensure that the scheduling strategy is implemented as actual green electricity consumption behavior, it is necessary to generate control commands that can be issued and executed based on the green electricity optimization scheduling scheme, and to schedule the green electricity power of various electrical equipment on the user side (including electric vehicle charging load, flexible load, adjustable equipment, etc.) in real time, accurately and orderly, so as to truly improve the green electricity consumption rate and ensure the safe and stable operation of the distribution network.

[0054] Specifically, firstly, based on the power allocation results of the green electricity optimization scheduling scheme, green electricity power scheduling instructions for individual electrical devices are generated according to information such as time granularity, device number, power limit, and operating period. These instructions include device identification, target green electricity power, execution time period, adjustment range, and allowable deviation range. Secondly, through communication methods such as the power Internet of Things, industrial Ethernet, power line carrier, 4G / 5G, or fieldbus, the green electricity power scheduling instructions are sent to the corresponding smart control cabinets, charging pile controllers, smart meters, building controllers, aggregation control terminals, or electric vehicle aggregation management platforms. Next, after receiving the scheduling instructions, each electrical device or its control unit adjusts its own power consumption according to the instructions, prioritizing the use of green electricity while ensuring that rigid electricity demand is not affected, and feeding back real-time operating power and status information to the scheduling unit. Finally, the scheduling unit collects the actual power consumption and green electricity usage of the devices in real time and compares them with the scheduling instructions; if a deviation occurs, rolling optimization is initiated based on the latest source-load data, and subsequent scheduling instructions are updated.

[0055] Preferably, green electricity dispatch instructions can also be converted into economic signals to incentivize electric vehicle users to respond to dispatch during periods of high GLMI (Gross Energy Management Index), and aggregators can generate dynamic green electricity trading prices linked to GLMI. The formula for dynamic green electricity trading prices is: in, The grid's regular time-of-use tariff or the green electricity market clearing price at time t; The maximum price discount is set by the aggregator based on marketing strategies and costs. Let t be the green electricity matching index at time t.

[0056] To address the aforementioned technical problems, another embodiment of the present invention provides a user-side green electricity optimization scheduling system. For details, please refer to [link to relevant documentation]. Figure 2 , Figure 2 The diagram shown illustrates the structure of a user-side green electricity optimization scheduling system according to one embodiment of the present invention. The system includes: The data acquisition module is used to acquire active power data of the power distribution network and power load data of various electrical devices; The feature vector module is used to obtain a source-side feature vector sequence based on the first time-series features of the active power data, and to obtain a load-side feature vector sequence based on the second time-series features of each of the electrical load data. The similarity analysis module is used to determine the source-load curve similarity information based on the source-side feature vector sequence and the load-side feature vector sequence. The green electricity matching index module is used to obtain the green electricity matching index based on the source-load curve similarity information and the dispatchable power information, wherein the dispatchable power information is determined by the operating reference data of the distribution network; The scheduling scheme module is used to construct a green electricity optimization scheduling model with the highest green electricity consumption rate as the objective function, and to determine the constraints of the green electricity optimization scheduling model based on green electricity supply constraints, dispatchable load constraints, total user electricity demand constraints, and distribution network security constraints. Based on the constraints and the green electricity matching index, the green electricity optimization scheduling model is solved to obtain the green electricity optimization scheduling scheme. The scheduling execution module is used to schedule the green power of each of the power-consuming devices on the user side according to the green power optimization scheduling scheme.

[0057] Compared with existing technologies, the beneficial effects of the embodiments of the present invention are at least one of the following: The present invention achieves deep extraction of multi-dimensional morphological features of the source-load curve by performing time-series feature analysis on the acquired active power data of the distribution network to obtain the source-side feature vector sequence, and performing time-series feature analysis on the load data of each electrical device to obtain the load-side feature vector sequence; The present invention achieves quantitative comparison of the green power output curve and the user-side load curve in terms of trend, fluctuation mode and local features by determining the similarity information of the source-load curve, thereby solving the technical defects of traditional Euclidean distance or simple correlation coefficient that cannot tolerate local scaling of the time axis and are difficult to accurately measure the consistency of source-load morphology; The present invention constructs a green power matching index based on the similarity information of the source-load curve and the dispatchable power information determined by the distribution network operation benchmark data, thereby integrating "source-load morphology matching potential" and "actual dispatchable capacity of the power grid" into a single comprehensive decision index, thereby solving the one-sidedness of dispatch decision caused by existing methods either ignoring the power grid carrying capacity or failing to quantify the degree of matching. The present invention addresses the following issues: It uses the highest green energy absorption rate as the objective function and simultaneously considers green energy supply constraints, dispatchable load constraints, total user electricity demand constraints, and distribution network security constraints to construct an optimized scheduling model. This achieves mathematical modeling for optimizing green energy utilization under multiple physical and user boundary conditions. Furthermore, by solving the green energy optimization scheduling model based on constraints and a green energy matching index, the present invention obtains an optimized green energy scheduling scheme. This strategy guides the priority allocation of green energy power during periods of high matching degree and high margin, thus solving the problems of low green energy absorption rate and wasted high green energy periods caused by the lack of spatiotemporal priority quantification indicators in existing technologies. Finally, the present invention schedules the green energy power of various user-side electrical equipment according to the optimized green energy scheduling scheme, achieving closed-loop linkage between source-load matching analysis and actual control commands. This solves the lag problem of traditional scheduling schemes, which only remain at the planning level and cannot be dynamically adjusted based on real-time operating baseline data. This improves the robustness and real-time executability of the green energy scheduling scheme, thereby contributing to ensuring the safe and stable operation of the distribution network.

[0058] The embodiments described above are merely illustrative of several implementations of the present invention, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of the present invention. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of the present invention, and these modifications and improvements all fall within the scope of protection of the present invention. Therefore, the scope of protection of this patent should be determined by the appended claims.

Claims

1. A user-side green electricity optimized scheduling method, characterized in that, include: Obtain active power data of the power distribution network and power load data of each electrical device; Based on the first time-series characteristics of the active power data, a source-side feature vector sequence is obtained, and based on the second time-series characteristics of each of the electricity load data, a load-side feature vector sequence is obtained. Based on the source-side feature vector sequence and the load-side feature vector sequence, the source-load curve similarity information is determined; Based on the source-load curve similarity information and the dispatchable power information, a green electricity matching index is obtained, wherein the dispatchable power information is determined by the operating baseline data of the distribution network; Using the highest green electricity absorption rate as the objective function, a green electricity optimization scheduling model is constructed. Based on green electricity supply constraints, dispatchable load constraints, total user electricity demand constraints, and distribution network security constraints, the constraints of the green electricity optimization scheduling model are determined. Based on the constraints and the green electricity matching index, the green electricity optimization scheduling model is solved to obtain the green electricity optimization scheduling scheme. According to the green electricity optimization scheduling scheme, the green electricity power of each of the user-side electrical devices is scheduled.

2. The user-side green electricity optimization scheduling method as described in claim 1, characterized in that, The first time-series feature based on the active power data yields a source-side feature vector sequence, including: The active power data is standardized to obtain standard active power data; Based on the aforementioned standard active power data, the source-side power curve is determined; The source-side power curve is processed to obtain the first timing feature; The first time-series feature is analyzed to obtain the source-side feature vector sequence.

3. The user-side green electricity optimization scheduling method as described in claim 2, characterized in that, The second time-series feature based on each of the electricity load data yields a load-side feature vector sequence, including: The power load data of each of the aforementioned electrical devices is standardized to obtain standard power load data. Based on all the aforementioned standard electricity load data, determine the load-side electricity load curve; The load curve on the load side is processed to obtain the second time-series characteristic; The second time-series feature is analyzed to obtain the load-side feature vector sequence.

4. The user-side green electricity optimization scheduling method as described in claim 3, characterized in that, The determination of source-load curve similarity information based on the source-side feature vector sequence and the load-side feature vector sequence includes: A first similarity analysis is performed on the source-side feature vector sequence and the load-side feature vector sequence to obtain first morphological similarity information; A second similarity analysis is performed on the source-side feature vector sequence and the load-side feature vector sequence to obtain second feature similarity information; A third similarity analysis is performed on the source-side power curve and the load-side electrical load curve to obtain third amplitude similarity information; The source-load curve similarity information is obtained by integrating the first morphological similarity information, the second feature similarity information, and the third amplitude similarity information.

5. The user-side green electricity optimization scheduling method as described in claim 4, characterized in that, The first similarity analysis of the source-side feature vector sequence and the load-side feature vector sequence to obtain first morphological similarity information includes: Based on the Euclidean distance calculation principle, each feature vector in the source-side feature vector sequence and each feature vector in the corresponding load-side feature vector sequence are processed to obtain the source-load feature distance matrix; The source-load feature distance matrix is ​​processed based on the dynamic time warping algorithm, and the processing result is analyzed to obtain the first morphological similarity information.

6. The user-side green electricity optimization scheduling method as described in claim 4, characterized in that, The second similarity analysis of the source-side feature vector sequence and the load-side feature vector sequence to obtain second feature similarity information includes: Based on the source-side feature vector sequence, a high-dimensional representation of the source-side feature vector is obtained, and based on the load-side feature vector sequence, a high-dimensional representation of the load-side feature vector is obtained. Based on the principle of cosine similarity, the high-dimensional representation results of the source-side feature vector and the high-dimensional representation results of the load-side feature vector are processed to obtain the second feature similarity information.

7. The user-side green electricity optimization scheduling method as described in claim 4, characterized in that, The third similarity analysis of the source-side power curve and the load-side electricity consumption curve yields third amplitude similarity information, including: The source-side power curve and the load-side electricity consumption curve are time-aligned respectively to obtain the corresponding source-side power curve and load-side electricity consumption curve to be analyzed. Based on the pre-constructed amplitude matching degree model, the power curve on the source side and the electricity load curve on the load side to be analyzed are processed to obtain the third amplitude similarity information.

8. The user-side green electricity optimization scheduling method as described in claim 3, characterized in that, The green electricity matching index is obtained based on the source-load curve similarity information and the schedulable power information, including: The load-side electrical load curve is analyzed to obtain load adjustability margin information; Based on the operating baseline data and the load adjustability margin information, the dispatchable power information is determined; The source-load curve similarity information and the schedulable power information are integrated to obtain the green electricity matching index.

9. The user-side green electricity optimization scheduling method as described in claim 1, characterized in that, The step of scheduling the green power of each of the user-side electrical devices according to the green power optimization scheduling scheme includes: Based on the green electricity optimization scheduling scheme, green electricity power scheduling instructions corresponding to each of the electrical devices are generated. According to the green power scheduling instruction, the green power of each of the electrical devices on the user side is scheduled.

10. A user-side green electricity optimization scheduling system, characterized in that, include: The data acquisition module is used to acquire active power data of the power distribution network and power load data of various electrical devices; The feature vector module is used to obtain a source-side feature vector sequence based on the first time-series features of the active power data, and to obtain a load-side feature vector sequence based on the second time-series features of each of the electrical load data. The similarity analysis module is used to determine the source-load curve similarity information based on the source-side feature vector sequence and the load-side feature vector sequence. The green electricity matching index module is used to obtain the green electricity matching index based on the source-load curve similarity information and the dispatchable power information, wherein the dispatchable power information is determined by the operating reference data of the distribution network; The scheduling scheme module is used to construct a green electricity optimization scheduling model with the highest green electricity consumption rate as the objective function, and to determine the constraints of the green electricity optimization scheduling model based on green electricity supply constraints, dispatchable load constraints, total user electricity demand constraints, and distribution network security constraints. Based on the constraints and the green electricity matching index, the green electricity optimization scheduling model is solved to obtain the green electricity optimization scheduling scheme. The scheduling execution module is used to schedule the green power of each of the power-consuming devices on the user side according to the green power optimization scheduling scheme.

Images