Power distribution management method and system based on high-proportion new energy admission

Abstract

The invention discloses a power distribution management method and system based on high-proportion new energy acceptance, and relates to the technical field of new energy, and the method comprises the steps: building a new energy consumption comprehensive index system according to an elastic coefficient index, an efficiency index, a new energy permeability index, an external system exchange index, an energy self-sufficiency index and a green and environmental protection index of a microgrid; constructing an input feature sequence according to the consumption comprehensive index score corresponding to the new energy consumption comprehensive index system, the regional voltage level of the microgrid, the new energy penetration ratio and the regional load level; constructing an output feature sequence according to the boundary condition, the maximum access amount and the access position of the distributed new energy; training a neural network according to the input feature sequence and the output feature sequence to obtain a prediction analysis model; and the prediction analysis model responds to the real-time input feature sequence to obtain a target output feature sequence, and performs micro-grid consumption based on the target output feature sequence, thereby overcoming the technical defects of insufficient safety and low operation efficiency of a traditional power distribution system.

Description

Technical Field

[0001] This invention relates to the field of new energy technology, specifically to a power distribution management method and system based on a high proportion of new energy integration. Background Technology

[0002] Microgrids, as small-scale energy systems integrating distributed power sources, loads, and distribution facilities, operate in both grid-connected and stand-alone modes. They are widely used in scenarios such as islands and remote areas, requiring energy storage systems to maintain power balance and efficiently utilize renewable energy. However, the intermittent and random nature of renewable energy sources currently leads to difficulties in grid connection and low absorption rates. With the integration of a high proportion of renewable energy, the distribution system adds a large number of new energy components, resulting in bidirectional power flow characteristics at nodes and branches, accompanied by significant uncertainties and power electronics phenomena. Traditional microgrid evaluation indicators for distribution systems cannot effectively assess the security of microgrids with a high proportion of renewable energy, leading to a lack of scientific planning for renewable energy absorption ratios in different scenarios, ultimately resulting in insufficient security and low operating efficiency of the distribution system.

[0003] The information disclosed in the background section is only intended to enhance the understanding of the background of this application, and therefore may include information that does not constitute prior art known to those skilled in the art. Summary of the Invention

[0004] The purpose of this invention is to address the core problems of insufficient power distribution system security and low operating efficiency caused by the intermittency and randomness of high-proportion renewable energy, which leads to difficulties in grid connection and low absorption rates. It proposes a power distribution management method and system based on the acceptance of high-proportion renewable energy. This method first constructs a comprehensive renewable energy absorption index system including elasticity coefficients and efficiency indices. Then, it acquires input-related data such as the voltage level of the microgrid area and output-related data such as the boundary conditions of distributed renewable energy. Furthermore, it constructs input and output feature sequences, trains a neural network to obtain a predictive analysis model, and finally obtains the target output based on the model's real-time input response to execute microgrid absorption. This overcomes the problems of ineffective traditional evaluation indicators, unscientific absorption ratio planning, and low system security and efficiency.

[0005] In a first aspect, one technical solution provided in this embodiment of the invention is a power distribution management method based on a high proportion of new energy source integration, comprising the following steps: A comprehensive indicator system for new energy consumption is constructed based on the microgrid's resilience coefficient, efficiency index, new energy penetration rate, external system exchange rate, energy self-sufficiency rate, and green environmental protection indicators. Obtain the regional voltage level, renewable energy penetration rate, regional load level, and boundary conditions, maximum capacity, and location of distributed renewable energy in the microgrid. The input feature sequence is constructed based on the comprehensive index score of renewable energy consumption, the regional voltage level of the microgrid, the renewable energy penetration ratio, and the regional load level; the output feature sequence is constructed based on the boundary conditions, maximum access capacity, and access location of distributed renewable energy. A predictive analysis model is obtained by training a neural network based on the input feature sequence and the output feature sequence. The predictive analysis model responds to the real-time input feature sequence to obtain the target output feature sequence, and performs microgrid absorption based on the target output feature sequence.

[0006] Preferably, the comprehensive index score corresponding to the comprehensive index system for new energy consumption is... The calculation formula is as follows: ; in, The comprehensive index score for the absorption of new energy microgrids is used. The elasticity coefficient index for new energy microgrids. The efficiency index of new energy microgrids. For green and environmentally friendly indicators of new energy microgrids, To increase the penetration rate of new energy sources, For external system exchange indicators of new energy microgrids, This refers to the energy self-sufficiency rate of new energy microgrids.

[0007] Preferably, the regional voltage levels are divided according to the AC voltage value of the current region. The AC voltage of the current region is greater than or equal to 110 kV and is determined to be the high voltage level. The AC voltage of the current region is less than or equal to 36 kV and greater than or equal to 1 kV and is determined to be the medium voltage level. The AC voltage of the current region is less than or equal to 220 V and is determined to be the low voltage level. The corresponding representative values ​​of the high voltage level, medium voltage level and low voltage level are marked as 1, 2 and 3.

[0008] Preferably, the boundary conditions include the voltage fluctuation range, frequency deviation, current fluctuation range, and frequency response time range of the access new energy source. The maximum access capacity refers to the maximum distributed renewable energy capacity that the current power system can accommodate. The access location refers to the operational node location of the current power system for introducing distributed new energy sources.

[0009] As a preferred option, the input feature value corresponding to the input feature sequence It is expressed as follows: ; in, This represents the voltage level of the region. The percentage of new energy penetration in the region. This represents the current regional load level.

[0010] Preferably, a predictive analysis model is obtained by training a neural network based on the input feature sequence and the output feature sequence, including the following steps: The target dataset is obtained by performing time-axis alignment verification on the input feature sequence and the output feature sequence through a time-domain alignment algorithm. The target dataset is then segmented to obtain the training dataset and the validation dataset. The neural network is trained using a training dataset to obtain an initial predictive analytics model, and the initial predictive analytics model is validated using a validation dataset to determine the target predictive analytics model.

[0011] Preferably, the step of performing time axis alignment verification on the input feature sequence and the output feature sequence using a time-domain alignment algorithm includes the following steps: Based on the Nyquist sampling theorem, the original sampling frequency distribution of the features corresponding to each timestamp is analyzed to determine the target sampling frequency. When the sampling frequency is lower than the target sampling frequency, a sliding window weighted average method is used for downsampling. When the sampling frequency is higher than the target sampling frequency, a cubic spline interpolation method is used for upsampling. The original data of the input feature sequence and the output feature sequence are obtained based on the target sampling frequency. Using the original data as samples, the cross-correlation coefficient between the input feature sequence and the output feature sequence is calculated to determine the lag time interval. The temporal morphology of the input and output features is aligned using the DTW algorithm. The sequence similarity under different lag time intervals is calculated with the minimum cumulative distance as the optimization objective, and the optimal lag parameter is determined for time axis alignment verification.

[0012] Secondly, an embodiment of the present invention also provides a technical solution: a power distribution management system based on high proportion of renewable energy acceptance, applicable to the above-mentioned power distribution management method based on high proportion of renewable energy acceptance, including: a comprehensive index construction module, a data acquisition module, a data integration module, a model construction module, and a absorption and modulation module; The comprehensive index construction module constructs a comprehensive index system for new energy consumption based on the microgrid's elasticity coefficient, efficiency index, new energy penetration rate, external system exchange index, energy self-sufficiency rate, and green environmental protection index. The data acquisition module acquires the regional voltage level, renewable energy penetration rate, regional load level, and boundary conditions, maximum access volume, and access location of the microgrid. The data integration module constructs an input feature sequence based on the comprehensive index score of the new energy consumption comprehensive index system, the regional voltage level of the microgrid, the new energy penetration ratio, and the regional load level; and constructs an output feature sequence based on the boundary conditions, maximum access volume, and access location of distributed new energy. The model building module trains a neural network based on the input feature sequence and the output feature sequence to obtain a predictive analysis model; The absorption modulation module obtains the target output characteristic sequence based on the real-time input characteristic sequence according to the predictive analysis model, and performs microgrid absorption based on the target output characteristic sequence.

[0013] Thirdly, one technical solution provided in this embodiment of the invention is: an electronic device, comprising: a processor and a memory; the memory is used to store a computer program, and when the processor executes the computer program, the electronic device executes the above-mentioned power distribution management method based on the high proportion of new energy acceptance.

[0014] Fourthly, one technical solution provided in this embodiment of the invention is a computer-readable storage medium, characterized in that it includes a computer program or instructions; when the computer program or instructions are run on a computer, the computer executes the above-mentioned power distribution management method based on the high proportion of new energy adoption.

[0015] The beneficial effects of this invention are: (1) In view of the fact that traditional microgrid evaluation indicators only cover a single operating dimension and cannot adapt to the bidirectional power flow, power electronics and uncertainty characteristics of the distribution system after a high proportion of new energy access, resulting in the failure of the safety assessment of high proportion of new energy microgrids, this application proposes a scheme to construct a multi-dimensional new energy consumption comprehensive indicator system. By selecting microgrid elasticity coefficient, efficiency index, new energy penetration rate, external system exchange index, energy self-sufficiency rate and green environmental protection index, the multi-dimensional indicators are synergistically coupled to form a complete evaluation system, so as to achieve a comprehensive and accurate characterization of the operating status of high proportion of new energy microgrids, break through the limitations of traditional single indicator assessment, provide scientific indicator basis for subsequent consumption decision-making, and thus solve the technical bottleneck that it is difficult to effectively quantify the safety of high proportion of new energy microgrids. (2) In response to the problem that the input data (regional voltage level, new energy penetration ratio, etc.) and output data (distributed new energy boundary conditions, maximum access volume, etc.) in high-proportion new energy microgrids are asynchronous in time and have large differences in data sampling frequency, resulting in low training data quality and insufficient prediction accuracy of neural network models, this application proposes a scheme to construct input and output feature sequences and train prediction analysis models in combination with time-domain alignment algorithms. First, input feature sequences are constructed based on comprehensive absorption index scores, regional voltage levels (divided into high voltage / medium voltage / low voltage levels and assigned values ​​according to 110kV and above / 1-36kV / 220V and below), new energy penetration ratio, etc. Output feature sequences are constructed based on distributed new energy boundary conditions (voltage fluctuation range, frequency deviation, etc.), maximum access volume and access location. Then, up / down sampling processing is performed based on the Nyquist sampling theorem, and the DTW algorithm is used to calculate the cross-correlation coefficient of the sequence and the optimal lag parameter to achieve time axis calibration, ensuring the consistency and effectiveness of the training data time sequence, thereby improving the mapping accuracy of the neural network model on the correlation between input parameters and key absorption parameters, and providing reliable model support for scientific planning of absorption strategies. (3) In view of the problem that the intermittent and random nature of renewable energy leads to rapid dynamic changes in the operating status of microgrids with a high proportion of renewable energy, and that traditional static absorption strategies cannot adapt to system changes in real time, resulting in low absorption efficiency and insufficient operational safety of the distribution system, this application proposes a scheme for dynamic absorption management based on a predictive analysis model. By using a trained neural network predictive analysis model, the model receives real-time input feature sequences such as the current regional voltage level, renewable energy penetration ratio, and regional load level of the microgrid. It quickly responds and outputs the corresponding target output feature sequences such as the boundary conditions of distributed renewable energy, the maximum access amount, and the access location. Based on the target output feature sequence, the microgrid absorption strategy is dynamically adjusted to achieve real-time adaptation to the fluctuations in renewable energy output, avoid system power imbalance caused by sudden changes in renewable energy output, and thus improve the absorption capacity of the distribution system for high proportion of renewable energy, while ensuring the operational safety and efficiency of the system.

[0016] The above description of the invention is merely an overview of the technical solution of the present invention. In order to better understand the technical means of the present invention and to implement it in accordance with the contents of the specification, and to make the above and other objects, features and advantages of the present invention more apparent and understandable, specific embodiments of the present invention are described below. Attached Figure Description

[0017] Other features, objects, and advantages of the invention will become more apparent from the following detailed description of non-limiting embodiments with reference to the accompanying drawings. The drawings are for illustrative purposes only and are not intended to limit the invention. Furthermore, the same reference numerals denote the same parts throughout the drawings.

[0018] Figure 1 This is a flowchart of the power distribution management method based on a high proportion of new energy sources according to the present invention.

[0019] Figure 2 This is a schematic diagram of the predictive analysis model of the present invention.

[0020] Figure 3 This is a structural block diagram of the power distribution management system based on a high proportion of new energy sources according to the present invention. Detailed Implementation

[0021] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only one preferred embodiment of this invention and are only used to explain this invention. They do not limit the scope of protection of this invention. All other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this invention.

[0022] Before discussing the exemplary embodiments in more detail, it should be mentioned that some exemplary embodiments are described as processes or methods depicted as flowcharts. Although the flowcharts describe the operations (or steps) as sequential processes, many of the operations (or steps) can be performed in parallel, concurrently, or simultaneously. Furthermore, the order of the operations can be rearranged. The process can be terminated when its operation is completed, but it may also have additional steps not included in the figures; the process may correspond to a method, function, procedure, subroutine, subroutine, etc.

[0023] Example 1: As Figure 1 As shown, the power distribution management method based on a high proportion of renewable energy integration includes the following steps: S1. Construct a comprehensive indicator system for new energy consumption based on the microgrid's resilience coefficient, efficiency index, new energy penetration rate, external system exchange rate, energy self-sufficiency rate, and green environmental protection indicators. S2. Obtain the regional voltage level, renewable energy penetration rate, regional load level, and boundary conditions, maximum access volume, and access location of the microgrid. S3. Construct an input feature sequence based on the comprehensive index score of renewable energy consumption, the regional voltage level of the microgrid, the renewable energy penetration rate, and the regional load level; construct an output feature sequence based on the boundary conditions, maximum access capacity, and access location of distributed renewable energy. S4. Train a neural network based on the input feature sequence and the output feature sequence to obtain a predictive analysis model; S5. The predictive analysis model responds to the real-time input feature sequence to obtain the target output feature sequence, and performs microgrid absorption based on the target output feature sequence.

[0024] In this embodiment, step S1 constructs a comprehensive indicator system for renewable energy consumption based on the microgrid's resilience coefficient, efficiency index, renewable energy penetration rate, external system exchange rate, energy self-sufficiency rate, and green environmental protection indicators. This achieves the technical effect of breaking through the limitations of traditional single-dimensional evaluation and providing a multi-dimensional and comprehensive quantitative assessment basis for the operation status of microgrids with a high proportion of renewable energy. Step S2 obtains the regional voltage level, renewable energy penetration rate, regional load level, and boundary conditions, maximum access volume, and access location of distributed renewable energy in the microgrid. This achieves the technical effect of providing a comprehensive and accurate data source for subsequent feature sequence construction and ensuring the reliability of the data foundation. Step S3 constructs an input feature sequence based on the comprehensive consumption indicator score, regional voltage level, renewable energy penetration rate, and regional load level, and based on the distributed renewable energy... The boundary conditions, maximum access capacity, and access location are used to construct the output feature sequence, which establishes the correlation dimension between microgrid operating state parameters and key absorption parameters, providing structured and targeted data support for neural network training. Step S4 trains the neural network based on the input and output feature sequences to obtain the predictive analysis model (including time-domain alignment algorithm for data processing), which achieves the technical effect of constructing a model that can accurately map the correlation between microgrid operating state and absorption parameters, solving the problem of insufficient model accuracy caused by data time series differences. Step S5 allows the predictive analysis model to respond to the real-time input feature sequence to obtain the target output feature sequence, and executes microgrid absorption based on this sequence, achieving the technical effect of dynamically adapting to the intermittent and random characteristics of renewable energy, adjusting the absorption strategy in real time, and ensuring the safe and efficient operation of the microgrid.

[0025] It should be noted that the resilience coefficient is defined as the ability of a microgrid to maintain core load stability and restore normal operation when faced with disturbances such as fluctuations in renewable energy output and sudden load changes. A higher value indicates that the system can more flexibly respond to renewable energy fluctuations. A larger resilience coefficient results in a larger overall absorption index for renewable energy microgrids, while a smaller resilience coefficient results in a smaller overall absorption index. Specifically: Resilience Coefficient = Core load guarantee rate / Disturbance response time, where the core load guarantee rate is the ratio of the actual power supply to the demand power supply during the disturbance (value 0-1), and the disturbance response time is the time from the occurrence of the disturbance to the restoration of normal operation. This ratio reflects the system's disturbance resistance and recovery capability.

[0026] Furthermore, the efficiency index is defined as a comprehensive reflection of the microgrid's efficiency in converting and utilizing renewable energy and its operational economy. It considers the efficiency of electricity supply and demand matching. A high efficiency index means the system can more effectively meet electricity demand. When the efficiency index changes, the comprehensive index of renewable energy consumption increases with the increase of the efficiency index and decreases with the decrease of the efficiency index. Specifically: Efficiency Index =a1·Actual power generation of new energy / Theoretical maximum power generation of new energy + a2·Unit revenue from new energy power generation / Unit operation and maintenance cost of new energy power generation, where a1 and a2 are weights (a1+a2=1, which can be set according to technical / economic priority. For example, when technology is prioritized, a1=0.8 and a2=0.2 are set). The former reflects energy conversion efficiency, and the latter reflects operational economy.

[0027] Furthermore, the renewable energy penetration rate is defined as the scale and contribution of renewable energy in a microgrid, reflecting the proportion of renewable energy. A higher penetration rate indicates a greater impact of renewable energy on the system. Specifically: Renewable Energy Penetration Rate = Total renewable energy generation of the microgrid during a specific period / Total electricity consumption of the microgrid during a specific period, with the result presented as a percentage.

[0028] Furthermore, the external system exchange index is defined as the energy exchange capability and coordination between the microgrid and the external main grid, reflecting the degree of dependence on the external grid. Specifically, it includes: external system exchange index. = Actual exchange volume between the microgrid and the external power grid / Maximum allowable exchange volume between the microgrid and the external power grid. The actual exchange volume is the sum of the absolute values ​​of electricity purchased and sold within a specific period. The maximum allowable exchange volume is determined by the grid dispatch. The closer the ratio is to 1, the more saturated the exchange capacity is.

[0029] Furthermore, the energy self-sufficiency rate is defined as the ability of a microgrid to meet its own load demand through internal distributed renewable energy and energy storage systems, reflecting its independent operation level. The higher the energy self-sufficiency rate, the lower the system's dependence on external energy sources. Specifically: Energy Self-Sufficiency Rate =(Power generation from new energy sources within the microgrid + Power discharge from energy storage systems) / Total power consumption of the microgrid. The result is presented as a percentage; the higher the value, the stronger the independent operation capability.

[0030] Furthermore, the green environmental protection index is defined as the degree of environmental friendliness of microgrid operation during the integration of new energy sources. It primarily reflects the emission reduction benefits of new energy replacing traditional energy sources. A higher green environmental protection index indicates that the system has a smaller environmental impact during the integration of new energy sources. The comprehensive index of new energy consumption increases with the increase of the green environmental protection index and decreases with the decrease of the green environmental protection index. Specifically: Green Environmental Protection Index = Microgrid renewable energy generation × (Carbon emission coefficient of traditional coal-fired power generation - Carbon emission coefficient of renewable energy generation) / Total microgrid power generation, where the carbon emission coefficient is in units of... This ratio reflects the emission reduction contribution per unit of electricity generated; the higher the value, the more significant the environmental benefits.

[0031] As an optional embodiment of this example, the comprehensive index score corresponding to the comprehensive index system for new energy consumption is... The calculation formula is as follows: ; in, The comprehensive index score for the absorption of new energy microgrids is used. The elasticity coefficient index for new energy microgrids. The efficiency index of new energy microgrids. For green and environmentally friendly indicators of new energy microgrids, To increase the penetration rate of new energy sources, For external system exchange indicators of new energy microgrids, This refers to the energy self-sufficiency rate of new energy microgrids.

[0032] As an optional embodiment of this example, the regional voltage levels are divided according to the AC voltage value of the current region. The AC voltage of the current region is greater than or equal to 110 kV and is determined to be the high voltage level. The AC voltage of the current region is less than or equal to 36 kV and greater than or equal to 1 kV and is determined to be the medium voltage level. The AC voltage of the current region is less than or equal to 220 V and is determined to be the low voltage level. The corresponding representative values ​​of the high voltage level, medium voltage level and low voltage level are marked as 1, 2 and 3.

[0033] In this embodiment, the impact of renewable energy generation on grid connection can be predicted more precisely based on different voltage levels. Different voltage levels have different power transmission and distribution characteristics, so taking them into account helps to predict the impact of renewable energy grid connection more accurately. Different voltage levels require different grid connection strategies and control measures. By distinguishing between high-voltage, medium-voltage, and low-voltage levels, the system can formulate targeted renewable energy grid connection schemes based on actual conditions to optimize the operation of the power system. The grid structure and operating characteristics of different voltage levels are different, especially in terms of transmission loss and stability. Incorporating voltage levels into the predictive analysis model helps to consider these factors more comprehensively, thereby better guiding the operation and management of renewable energy microgrids. By using representative values ​​of voltage levels as one of the features, the model's ability to distinguish between different situations can be increased, thereby improving the accuracy of prediction. Different voltage levels have different impacts on renewable energy grid connection, and incorporating them into the model can better capture these impacts. Dividing voltage levels into representative values ​​helps with data integration and standardization. Voltage standards vary in different regions. By abstracting voltage levels into several representative values, the predictive analysis model can be made more universal and applicable to applications in different regions.

[0034] As an optional embodiment of this example, the boundary conditions include the voltage fluctuation range of the access new energy source, the frequency deviation, the current fluctuation range, and the frequency response time range. The maximum access capacity refers to the maximum distributed renewable energy capacity that the current power system can accommodate. The access location refers to the operational node location of the current power system for introducing distributed new energy sources.

[0035] It is understood that this embodiment employs explicit boundary conditions, specifically including the voltage fluctuation range, frequency deviation, current fluctuation range, and frequency response time range of the accessed renewable energy. Simultaneously, it defines the maximum access quantity as the maximum distributed renewable energy capacity that the current power system can accommodate, and the access location as the operating node location of the power system for introducing distributed renewable energy. By synergistically linking the four key parameters of the boundary conditions with the maximum access quantity and access location, the voltage, frequency, current, and response time parameters of the boundary conditions limit the operating threshold for distributed renewable energy access. The maximum access quantity clarifies the upper limit of the system's capacity, and the access location determines the physical node for renewable energy access. These three elements are coupled to form a complete constraint system of operating threshold - capacity upper limit - physical node. This achieves comprehensive coordinated control from parameter constraints and capacity management to node positioning. It avoids system stability issues caused by excessive voltage / frequency fluctuations and untimely responses, prevents power imbalances caused by renewable energy access exceeding the system's carrying capacity, and reduces power transmission losses through precise node positioning. Ultimately, this ensures the safe and efficient operation of distributed renewable energy in the power system.

[0036] As an optional embodiment of this example, the input feature value corresponding to the input feature sequence is... It is expressed as follows: ; in, This represents the voltage level of the region. The percentage of new energy penetration in the region. This represents the current regional load level.

[0037] Understandably, the comprehensive absorption index reflects the overall impact of renewable energy generation on the power system. An increase in the comprehensive absorption index indicates a greater impact from renewable energy, leading to changes in the stability and efficiency of the power system. A higher comprehensive absorption index means the system needs better adaptive measures to maintain stable operation. The voltage level represents the voltage level of the power system in a given region. Different voltage levels have different impacts on renewable energy integration. Higher voltage levels indicate a greater carrying capacity of the power system, making it easier to accommodate renewable energy. Considering city or community-level power systems with lower voltages, this is suitable for the integration of distributed energy sources, such as solar photovoltaic systems. The focus is on urban or regional power grids, involving large-scale distributed energy integration, such as wind power and small-scale biomass power generation. Research is conducted on the power system backbone, including large-scale distributed energy sources such as large wind farms and solar power plants. The renewable energy penetration rate represents the proportion of renewable energy in total power generation. As the renewable energy penetration rate increases, the energy composition of the power system changes, affecting its stability and supply-demand matching. A high renewable energy penetration rate requires more refined scheduling and management. The current load level reflects the current load demand of the power system. Higher load levels require more energy supply to meet demand, and the integration of renewable energy will impact the supply-demand balance.

[0038] As an optional embodiment of this example, such as Figure 2 As shown, the predictive analysis model is obtained by training a neural network based on the input feature sequence and the output feature sequence, including the following steps: The target dataset is obtained by performing time-axis alignment verification on the input feature sequence and the output feature sequence through a time-domain alignment algorithm. The target dataset is then segmented to obtain the training dataset and the validation dataset. The neural network is trained using a training dataset to obtain an initial predictive analytics model, and the initial predictive analytics model is validated using a validation dataset to determine the target predictive analytics model.

[0039] It is understood that this embodiment employs a technique of first using a time-domain alignment algorithm to perform time-axis alignment verification on the input and output feature sequences to obtain the target dataset, then splitting the target dataset into a training dataset and a validation dataset (e.g., a 7:3 ratio between the training and validation datasets), subsequently using the training dataset to train a neural network to obtain a nascent predictive analysis model, and finally using the validation dataset to validate the nascent predictive analysis model to determine the target predictive analysis model. The time-domain alignment algorithm eliminates temporal misalignment caused by differences in sampling frequencies between the input and output feature sequences, providing a time-consistent and reliable target dataset for subsequent model training. Splitting the target dataset into training and validation datasets allows the training dataset to be used for iterative optimization of neural network parameters to build a nascent model with basic predictive capabilities, while the validation dataset is used to test the generalization and prediction accuracy of the nascent model. The two work together to achieve data calibration and model training, ensuring that the target predictive analysis model can accurately map the correlation between input and output features, avoiding prediction errors caused by data temporal deviations or model overfitting / underfitting, and providing reliable model support for the dynamic execution of subsequent microgrid absorption strategies.

[0040] As an optional embodiment of this example, the step of performing time axis alignment verification on the input feature sequence and the output feature sequence using a time-domain alignment algorithm includes the following steps: Based on the Nyquist sampling theorem, the original sampling frequency distribution of the features corresponding to each timestamp is analyzed to determine the target sampling frequency. When the sampling frequency is lower than the target sampling frequency, a sliding window weighted average method is used for downsampling. When the sampling frequency is higher than the target sampling frequency, a cubic spline interpolation method is used for upsampling. The original data of the input feature sequence and the output feature sequence are obtained based on the target sampling frequency. Using the original data as samples, the cross-correlation coefficient between the input feature sequence and the output feature sequence is calculated to determine the lag time interval. The temporal morphology of the input and output features is aligned using the DTW algorithm. The sequence similarity under different lag time intervals is calculated with the minimum cumulative distance as the optimization objective, and the optimal lag parameter is determined for time axis alignment verification.

[0041] Understandably, this embodiment uses the Nyquist sampling theorem to provide a scientific basis for adjusting the sampling frequency, ensuring that data sampling meets the requirement of no distortion. The sliding window weighted average method and the cubic spline interpolation method are adapted to different frequency deviation scenarios, ensuring the integrity and accuracy of the data after the sampling frequency is adjusted. The cross-correlation coefficient calculation provides quantitative support for determining the lag time interval, while the DTW algorithm accurately matches the temporal pattern of the input and output sequences. The two work together to optimize the lag parameters, ultimately eliminating the temporal misalignment problem of the input and output feature sequences. This provides a temporally consistent and reliable dataset for subsequent neural network model training, effectively improving the accuracy of the model's mapping of the correlation between input and output features.

[0042] For example, in the power distribution management method based on high proportion of renewable energy in this embodiment, to address the issue of inconsistent sampling frequencies between the input feature sequence (including comprehensive absorption index score, regional voltage level, etc.) and the output feature sequence (including distributed renewable energy boundary conditions, maximum access volume, etc.), frequency analysis is first conducted based on the Nyquist sampling theorem: First, the original sampling frequencies of the input and output features at each timestamp are statistically analyzed (e.g., regional voltage level data is sampled once every 5 minutes, and renewable energy penetration ratio data is sampled once every 10 minutes). According to the theorem that "the target sampling frequency must be no less than twice the highest frequency of the signal", a unified target sampling frequency is determined by combining the highest original sampling frequency in the two types of sequences (e.g., if the highest original frequency is once every 5 minutes, then the target sampling frequency is set to once every 2.5 minutes).

[0043] Next, frequency normalization is performed: if the original sampling frequency of a certain feature is higher than the target sampling frequency (e.g., the new energy penetration ratio is once every 10 minutes, which is lower than the target of once every 2.5 minutes), cubic spline interpolation is used to supplement the missing timestamp data by fitting the function curve of adjacent sampling points to achieve upsampling; if the original sampling frequency is lower than the target sampling frequency (e.g., a certain load data is once every 1 minute, which is higher than the target of once every 2.5 minutes), the sliding window weighted average method is used, with the time interval corresponding to the target sampling frequency as the window, and the average value of multiple sampling values ​​in the window is calculated according to weight to achieve downsampling. Finally, based on the target sampling frequency, the original input and output feature data with unified time series are integrated to obtain the original data.

[0044] Then, the lag time interval is determined: using the normalized original data as a sample, assuming the input sequence is... The output sequence is Where n is the number of samples in the original data, the cross-correlation coefficient between the two sequences is calculated. , , The mean of the input feature sequence X; , To output the mean of the feature sequence Y, the strength of the linear correlation between the two sequences is quantified by calculating the ratio of the product of the covariance of the two sequences' deviations from their respective means to the product of their standard deviations. The value ranges from [-1, 1]. The closer it is to 1 or -1, the stronger the linear correlation between the two sequences. The value is calculated using sliding motion with different lag steps. This allows us to determine the lag time interval with the strongest correlation, providing a basis for subsequent DTW algorithm timing calibration. For example, by sliding and trying different lag step sizes (such as lag by 1 or 2 target sampling periods), the range of step sizes with the largest absolute value of the cross-correlation coefficient is determined as the lag time interval (for example, the strongest correlation occurs when lag by 1-3 sampling periods).

[0045] Finally, the DTW algorithm is used to complete the temporal alignment: based on the lag time interval, a temporal morphology matrix of the input and output feature sequences is constructed. Taking the minimum cumulative distance between corresponding points of the two sequences as the optimization objective, the sequence similarity under different lag parameters is calculated (such as the minimum cumulative distance when the lag is 2 sampling periods). This parameter is determined as the optimal lag parameter, and the time axis of the input and output feature sequences is calibrated accordingly. Finally, the time axis alignment verification is completed to ensure that the two types of sequences are accurately matched in the time dimension, providing high-quality data for subsequent neural network model training.

[0046] Specifically, the calculation process for timing alignment using the DTW algorithm is as follows: Calculate the element-wise distances between the input sequence and the output sequence under different lag parameters, and construct a distance matrix. , ,in, This represents the Euclidean distance between the i-th element of the input sequence and the j-th element of the output sequence after a lag of k steps. Calculate the cumulative distance matrix based on distance matrix D. ,satisfy: ; The initial conditions are: And when i=0 or j=0, At this point, there are invalid boundary values. For each lag parameter k, the cumulative distance (i.e., sequence similarity metric) between the input and output sequences is: Optimal lag parameter The lag value that minimizes the cumulative distance: ; Through the above calculation process, the DTW algorithm can find the optimal lag parameter that makes the temporal morphology of the input feature sequence and the output feature sequence most similar (with the smallest cumulative distance) within different lag time intervals, thereby achieving time axis alignment verification.

[0047] Example 2: Another technical solution provided in this embodiment of the invention is a power distribution management system based on a high proportion of renewable energy adoption, applicable to the aforementioned power distribution management method based on a high proportion of renewable energy adoption, such as... Figure 3 As shown, it includes: a comprehensive index construction module 101, a data acquisition module 102, a data integration module 103, a model construction module 104, and an absorption modulation module 105; The comprehensive index construction module constructs a comprehensive index system for new energy consumption based on the microgrid's elasticity coefficient, efficiency index, new energy penetration rate, external system exchange index, energy self-sufficiency rate, and green environmental protection index. The data acquisition module acquires the regional voltage level, renewable energy penetration rate, regional load level, and boundary conditions, maximum access volume, and access location of the microgrid. The data integration module constructs an input feature sequence based on the comprehensive index score of the new energy consumption comprehensive index system, the regional voltage level of the microgrid, the new energy penetration ratio, and the regional load level; and constructs an output feature sequence based on the boundary conditions, maximum access volume, and access location of distributed new energy. The model building module trains a neural network based on the input feature sequence and the output feature sequence to obtain a predictive analysis model; The absorption modulation module obtains the target output characteristic sequence based on the real-time input characteristic sequence according to the predictive analysis model, and performs microgrid absorption based on the target output characteristic sequence.

[0048] It is understood that this embodiment adopts a system architecture adapted to the distribution management method for high-proportion renewable energy acceptance, consisting of a comprehensive index construction module, a data acquisition module, a data integration module, a model construction module, and an absorption modulation module. The comprehensive index construction module constructs a comprehensive index system for renewable energy absorption based on six types of indicators, including microgrid elasticity coefficient and efficiency index, providing a multi-dimensional evaluation benchmark for the system. The data acquisition module simultaneously acquires operational data such as regional voltage level and renewable energy penetration ratio, as well as absorption parameters such as distributed renewable energy boundary conditions, providing a complete data source for subsequent processing. The data integration module integrates the comprehensive index score and operational data into an input feature sequence and integrates the absorption parameters into an output feature sequence, establishing a correlation between the evaluation benchmark, operational data, and absorption parameters. The model construction module trains a neural network based on the two types of feature sequences to obtain a predictive analysis model, realizing the correlation mapping between the operating state and the absorption strategy. The absorption modulation module relies on the model response to input the feature sequence in real time, output the target absorption sequence, and execute the absorption operation. All modules work together to achieve systematic control of the entire process of high-proportion renewable energy microgrid absorption, ensuring the safety and efficiency of system operation.

[0049] Example 3: An optional embodiment provided in this invention is: an electronic device, including a memory and a processor, wherein the memory stores a computer program, and the processor calls the computer program in the memory to implement the steps of a power distribution management method based on a high proportion of new energy acceptance.

[0050] Example 4: An optional embodiment provided in this invention is: a storage medium storing computer-executable instructions, wherein when the computer-executable instructions are loaded and executed by a processor, the steps of a power distribution management method based on a high proportion of new energy adoption are implemented.

[0051] Through the above description of the embodiments, those skilled in the art will understand that, for the sake of convenience and brevity, only the division of the above functional modules is used as an example. In actual applications, the above functions can be assigned to different functional modules as needed, that is, the internal structure of the specific device can be divided into different functional modules to complete all or part of the functions described above.

[0052] In the embodiments provided in this application, it should be understood that the disclosed structures and methods can be implemented in other ways. For example, the structural embodiments described above are merely illustrative. For instance, the division of modules or units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another structure, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces, or indirect coupling or communication connection between structures or units, and may be electrical, mechanical, or other forms.

[0053] The units described as separate components may or may not be physically separate. A component shown as a unit can be one or more physical units; that is, it can be located in one place or distributed in multiple different locations. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0054] Furthermore, in the embodiments of this application, the functional units can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0055] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a readable storage medium. Based on this understanding, the technical solutions of the embodiments of this application, in essence, or the parts that contribute to the prior art, or all or part of the technical solutions, can be embodied in the form of a software product. This software product is stored in a storage medium and includes several instructions to cause a device (which may be a microcontroller, chip, etc.) or processor to execute all or part of the steps of the methods of the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0056] The specific embodiments described above are preferred embodiments of the power distribution management method and system based on high proportion of renewable energy adoption of the present invention, and are not intended to limit the specific scope of the present invention. The scope of the present invention includes but is not limited to the specific embodiments described above. All equivalent changes made in accordance with the shape and structure of the present invention are within the protection scope of the present invention.

Claims

1. A power distribution management method based on a high proportion of renewable energy integration, characterized in that: Includes the following steps: A comprehensive indicator system for new energy consumption is constructed based on the microgrid's resilience coefficient, efficiency index, new energy penetration rate, external system exchange rate, energy self-sufficiency rate, and green environmental protection indicators. Obtain the regional voltage level, renewable energy penetration rate, regional load level, and boundary conditions, maximum capacity, and location of distributed renewable energy in the microgrid. The input feature sequence is constructed based on the comprehensive index score of the new energy consumption comprehensive index system, the regional voltage level of the microgrid, the new energy penetration ratio and the regional load level. The output feature sequence is constructed based on the boundary conditions, maximum access volume, and access location of distributed new energy sources. A predictive analysis model is obtained by training a neural network based on the input feature sequence and the output feature sequence. The predictive analysis model responds to the real-time input feature sequence to obtain the target output feature sequence, and performs microgrid absorption based on the target output feature sequence.

2. The power distribution management method based on high proportion of new energy source integration according to claim 1, characterized in that: The comprehensive index score corresponding to the comprehensive index system for new energy consumption The calculation formula is as follows: ； in, The comprehensive index score for the absorption of new energy microgrids is used. The elasticity coefficient index for new energy microgrids. The efficiency index of new energy microgrids. For green and environmentally friendly indicators of new energy microgrids, To increase the penetration rate of new energy sources, For external system exchange indicators of new energy microgrids, This refers to the energy self-sufficiency rate of new energy microgrids.

3. The power distribution management method based on high proportion of new energy source integration according to claim 1, characterized in that: The regional voltage levels are divided according to the current AC voltage value of the region. The current region's AC voltage is greater than or equal to 110 kV and is determined to be a high voltage level. The current region's AC voltage is less than or equal to 36 kV and greater than or equal to 1 kV and is determined to be a medium voltage level. The current region's AC voltage is less than or equal to 220 V and is determined to be a low voltage level. The corresponding representative values ​​of the high voltage level, medium voltage level and low voltage level are marked as 1, 2 and 3.

4. The power distribution management method based on high proportion of new energy source integration according to claim 1, characterized in that: The boundary conditions include the voltage fluctuation range, frequency deviation, current fluctuation range, and frequency response time range of the access new energy source. The maximum access capacity refers to the maximum distributed renewable energy capacity that the current power system can accommodate. The access location is the operational node location for the introduction of distributed new energy sources into the current power system.

5. The power distribution management method based on high proportion of new energy source integration according to claim 1, characterized in that: Input feature values ​​corresponding to the input feature sequence It is expressed as follows: ； in, This represents the voltage level of the region. The percentage of new energy penetration in the region. This represents the current regional load level.

6. The power distribution management method based on high proportion of new energy source integration according to claim 1, characterized in that: The predictive analysis model is obtained by training a neural network based on the input feature sequence and the output feature sequence, including the following steps: The target dataset is obtained by performing time-axis alignment verification on the input feature sequence and the output feature sequence through a time-domain alignment algorithm. The target dataset is then segmented to obtain the training dataset and the validation dataset. The neural network is trained using a training dataset to obtain an initial predictive analytics model, and the initial predictive analytics model is validated using a validation dataset to determine the target predictive analytics model.

7. The power distribution management method based on high proportion of new energy source integration according to claim 6, characterized in that: The step of performing time-axis alignment verification on the input and output feature sequences using a time-domain alignment algorithm includes the following steps: Based on the Nyquist sampling theorem, the original sampling frequency distribution of the features corresponding to each timestamp is analyzed to determine the target sampling frequency. When the sampling frequency is lower than the target sampling frequency, a sliding window weighted average method is used for downsampling. When the sampling frequency is higher than the target sampling frequency, a cubic spline interpolation method is used for upsampling. The original data of the input feature sequence and the output feature sequence are obtained based on the target sampling frequency. Using the original data as samples, the cross-correlation coefficient between the input feature sequence and the output feature sequence is calculated to determine the lag time interval. The temporal morphology of the input and output features is aligned using the DTW algorithm. The sequence similarity under different lag time intervals is calculated with the minimum cumulative distance as the optimization objective, and the optimal lag parameter is determined for time axis alignment verification.

8. A distribution management system based on high-proportion renewable energy integration, applicable to the distribution management method based on high-proportion renewable energy integration as described in claims 1-7, characterized in that, include: The system comprises a comprehensive indicator construction module, a data acquisition module, a data integration module, a model construction module, and a damping and modulation module. The comprehensive index construction module constructs a comprehensive index system for new energy consumption based on the microgrid's elasticity coefficient, efficiency index, new energy penetration rate, external system exchange index, energy self-sufficiency rate, and green environmental protection index. The data acquisition module acquires the regional voltage level, renewable energy penetration rate, regional load level, and boundary conditions, maximum access volume, and access location of the microgrid. The data integration module constructs an input feature sequence based on the comprehensive index score of the new energy consumption comprehensive index system, the regional voltage level of the microgrid, the new energy penetration ratio, and the regional load level. The output feature sequence is constructed based on the boundary conditions, maximum access volume, and access location of distributed new energy sources. The model building module trains a neural network based on the input feature sequence and the output feature sequence to obtain a predictive analysis model; The absorption modulation module obtains the target output characteristic sequence based on the real-time input characteristic sequence according to the predictive analysis model, and performs microgrid absorption based on the target output characteristic sequence.

9. An electronic device, characterized in that, include: Processor and memory; The memory is used to store a computer program, which, when executed by the processor, causes the electronic device to perform the power distribution management method based on the high proportion of new energy adoption as described in any one of claims 1-7.

10. A computer-readable storage medium, characterized in that, include: A computer program or instruction; when the computer program or instruction is run on a computer, it causes the computer to perform the power distribution management method based on the high proportion of new energy sources as described in any one of claims 1-7.

Images