Multi-source sensing power load distribution platform

Abstract

The invention discloses a multi-source sensing power load distribution platform, and relates to the technical field of power distribution, and the platform comprises a forward standardization processing module which is used for determining power sensing data; the condition initialization module is used for determining a target uncertain condition; the model training module is used for supervising and training a load distribution model; and the power load distribution management module is used for performing power load distribution management on the local power grid. The technical problems that the load distribution optimization effect is poor and resource utilization is unbalanced due to the fact that uncertainty factors of the power equipment and the load end are not considered in the power load distribution process are solved, the real-time power load distribution optimization of the multi-source sensing data is achieved so as to adapt to different load changes and power requirements, and the real-time power load distribution optimization of the multi-source sensing data is achieved. The power load distribution efficiency is improved; the power line loss is reduced; and the operation stability of a local power grid is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of power distribution, and in particular to a multi-source-aware power load distribution platform. Background Art

[0002] With the transformation of the global energy structure and the continuous development of smart grid technology, the power system is facing unprecedented challenges. In the traditional power distribution process, there are problems such as single data source, delayed information transmission, and unbalanced load distribution, which lead to low efficiency of power resource utilization, poor power supply security, and inability to effectively cope with the uncertainty brought by large-scale new energy access. Especially in local power grids, due to the complex influence of multiple factors such as power equipment capacity, load characteristics, and local power grid topology, the dynamic adjustment of power load distribution becomes particularly difficult. At present, most power load distribution strategies rely on traditional scheduling methods based on fixed parameters and rules, lack effective integration and analysis of multi-source data, and fail to fully consider the uncertainty factors of power equipment and load ends, resulting in low load scheduling efficiency, poor grid operation stability, and the rational allocation of power resources cannot be effectively guaranteed. Summary of the invention

[0003] The present application provides a multi-source-aware power load distribution platform, which is used to solve the technical problems of poor load distribution optimization effect and unbalanced resource utilization due to failure to consider uncertainty factors between power equipment and load ends during the power load distribution process.

[0004] The present application provides a multi-source-aware power load distribution platform, the platform comprising: A forward normalization processing module is used to connect a multi-source data interface and determine power perception data, wherein the power perception data includes first power sampling data determined by calling multi-source data related to source and load and performing downsampling fusion, and second power index data determined by performing forward normalization processing on the power index; a condition initialization module is used to introduce source-load uncertainty conditions, traverse the local power grid topology, perform condition initialization based on the production capacity specificity of the power equipment end and the energy consumption specificity based on the load end, and determine the target uncertainty conditions; a model training module is used to supervise the training of a load distribution model in combination with the local power grid topology and the target uncertainty conditions, wherein the local power grid topology is marked with power line loss, and load distribution optimization training is performed by lightweight processing and rasterization processing based on expansion and causality; a power load distribution management module is used to transmit the power perception data to the load distribution model, make decisions to determine the power load distribution strategy, and perform power load distribution management on the local power grid.

[0005] In a possible implementation, the model training module also performs the following processing: calling load distribution samples to supervise the training of the large distribution model with the local power grid topology, the load distribution samples including power sample data-uncertain condition samples-power distribution samples; based on the preset loss degree, lightweight processing is performed on the large distribution model, dividing the expansion grid and the causal grid, and supervised training of the load distribution model.

[0006] In a possible implementation, the model training module also performs the following processing: based on the preset loss degree, determining the lightweight transfer probability, wherein the lightweight transfer probability is the logical transfer probability of the elements of the large allocation model; dividing the expansion grid based on the preset scale characteristics under the time series and spatial distribution, wherein each grid causality has the same expansion standard; based on the power causal relationship, dividing the causal grid, wherein each causal grid has the same power causal relationship; combining the lightweight transfer probability, the expansion grid and the causal grid, the large allocation model is supervised lightweight supervised training to determine the load distribution model.

[0007] In a possible implementation, the model training module further performs the following processing: obtaining a first load distribution sample, transmitting it to the distribution large model, determining a first soft sample, wherein the first soft sample includes a distribution logic chain; traversing the distribution logic chain, performing link node identification and grid marking based on an expansion grid and a causal grid, and determining a first marked soft sample; and performing a lightweight training based on the lightweight transfer probability and combining the first load distribution sample with the first soft sample.

[0008] In a possible implementation, the power load distribution management module also performs the following processing: taking multi-energy mutual assistance as the first load distribution condition, wherein the multi-energy distribution priority is used as a constraint; taking multi-microgrid mutual assistance as the second load distribution condition, wherein the local distribution priority is used as a constraint; based on the first load distribution condition and the second load distribution condition, assisting in power load distribution decision-making.

[0009] In a possible implementation, the forward standardization processing module also performs the following processing: mining power indicators of power load distribution, wherein the power indicators are marked with weight values; traversing the power indicators, performing indicator type attribution and indicator attribute forward processing, and determining extremely large power indicators, wherein the indicator types include extremely large, extremely small, and interval types; matrixing and integrating the extremely large power indicators and performing standardization processing to determine the second power indicator data.

[0010] In a possible implementation, the forward normalization processing module further performs the following processing: traversing the multi-source data, performing a first downsampling process of the data granularity based on the data correlation, and determining the first downsampling data; traversing the first downsampling data, performing a second downsampling process based on the data dimension, and determining the second downsampling data; performing homologous data fusion on the second downsampling data to determine the first power sampling data.

[0011] In a possible implementation, the power load distribution management module also performs the following processing: tracking the response of the power load distribution strategy to determine the distribution response data; traversing the distribution response data to verify and trace the abnormal distribution, and determining the cause of the abnormal distribution, wherein the deviation based on the distribution volume and the deviation based on the distribution trend are used as verification criteria; and based on the abnormal distribution cause, distribution feedback management of the local power grid is performed.

[0012] One or more technical solutions provided in this application have at least the following technical effects or advantages: The present application connects a multi-source data interface to determine power perception data, wherein the power perception data includes first power sampling data determined by calling multi-source data related to source and load and performing downsampling fusion, and second power index data determined by performing forward normalization processing on the power index; introducing source-load uncertainty conditions, traversing the local power grid topology, performing conditional initialization based on the production capacity specificity of the power equipment end and the energy consumption specificity based on the load end, and determining the target uncertainty conditions; combining the local power grid topology with the target uncertainty conditions, supervising the training of the load distribution model, wherein the local power grid topology is marked with power line loss, and optimizing the load distribution training by lightweight processing and rasterization processing based on expansion and causality; transmitting the power perception data to the load distribution model, making decisions to determine the power load distribution strategy, and performing power load distribution management on the local power grid. The present invention solves the technical problem that the uncertainty factors of power equipment and load ends are not taken into consideration during the power load distribution process, resulting in poor load distribution optimization effect and unbalanced resource utilization. Through real-time power load distribution optimization based on multi-source sensing data, it can adapt to different load changes and power demands, thereby achieving the technical effect of improving power load distribution efficiency, reducing power line losses and enhancing the operating stability of the local power grid. BRIEF DESCRIPTION OF THE DRAWINGS

[0013] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0014] Figure 1 A schematic diagram of the structure of a multi-source sensing power load distribution platform provided in an embodiment of the present application; Figure 2 A flow chart of a training load distribution model for a multi-source-aware power load distribution platform provided in an embodiment of the present application.

[0015] Explanation of the reference numerals: forward normalization processing module 11 , condition initialization module 12 , model training module 13 , power load distribution management module 14 . DETAILED DESCRIPTION

[0016] The present application provides a multi-source perception power load distribution platform to solve the technical problems of poor load distribution optimization effect and unbalanced resource utilization due to failure to consider the uncertainty factors of power equipment and load ends in the power load distribution process. Through real-time power load distribution optimization based on multi-source perception data, it can adapt to different load changes and power demands, thereby achieving the technical effect of improving power load distribution efficiency, reducing power line losses and enhancing the operating stability of local power grids.

[0017] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of this application.

[0018] It should be noted that any variations of the terms "include" and "have" are intended to cover non-exclusive inclusions. For example, a process, method, platform, product or server that includes a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or modules that are not explicitly listed or inherent to these processes, methods, products or devices.

[0019] Examples, such as Figure 1 As shown, the embodiment of the present application provides a multi-source-aware power load distribution platform, which includes: The forward normalization processing module 11 is used to connect the multi-source data interface to determine the power sensing data, wherein the power sensing data includes the first power sampling data determined by calling the multi-source data related to the source and load and performing downsampling fusion, and the second power indicator data determined by performing forward normalization processing on the power indicator.

[0020] In an embodiment of the present application, a forward normalization processing module is used to connect a multi-source data interface to generate power sensing data. The power sensing data consists of two parts: first power sampling data and second power index data. One is the first power sampling data generated by calling multi-source data related to source and load and undergoing two-stage downsampling and homologous data fusion processing; the other is the second power index data generated by mining, type classification, forward processing and standardized matrix processing of power indicators. These two parts together constitute the power sensing data.

[0021] Furthermore, in the platform provided in the application embodiment, the forward normalization processing module 11 is also used for: The multi-source data are traversed, and based on the data correlation, a first downsampling process of the data granularity is performed to determine the first downsampling data; the first downsampling data are traversed, and a second downsampling process is performed based on the data dimension to determine the second downsampling data; the second downsampling data is fused with the same source data to determine the first power sampling data.

[0022] In the embodiment of the present application, multi-source data is first traversed and the first downsampling process is performed based on the correlation. At this stage, multi-source data is collected, including the capacity data of power generation equipment (such as real-time power generation), the load data of power consumption equipment (such as instantaneous load), and the external environment data (such as temperature, humidity, and wind speed). After preprocessing these data, cleaning abnormal values ​​(such as negative power generation) and deduplication, the Pearson correlation coefficient is used to calculate the correlation between each data and the target variable (such as load change or power line loss). Data with high correlation usually have a strong statistical connection. When the Pearson correlation coefficient is close to 1 (positive correlation) or -1 (negative correlation), it is considered to be highly correlated. For example, the correlation coefficient between power generation and load change is 0.85, which is considered to be highly correlated; when it is close to 0, such as the correlation coefficient between ambient humidity and line loss is 0.1, it is considered to be low correlation and can be eliminated. In this process, the data granularity is based on the original acquisition frequency (seconds, minutes, hours), and the first downsampling data is formed by screening the key data sets with a correlation threshold greater than 0.5, covering the key operating characteristics of the power system.

[0023] Then traverse the first down-sampled data and perform the second down-sampled processing based on the principal component analysis (PCA) method. Organize the first down-sampled data into a feature matrix, where each column represents a feature dimension, such as power generation, load rate, line loss rate, etc. The variance contribution rate of each feature dimension is calculated using the PCA algorithm. For example, the contribution rate of power generation is 40%, the contribution rate of load rate is 30%, the contribution rate of line loss rate is 20%, and the contribution rate of ambient temperature is only 5%. Set a threshold based on the cumulative variance contribution rate (such as 90%), select the feature dimension with the highest contribution rate, and eliminate the dimensions that contribute less to the overall information (such as ambient temperature). This process extracts the main features that have the greatest impact on power load distribution, reduces dimensional redundancy, and generates more refined second down-sampled data.

[0024] Finally, the second down-sampled data is fused with the same source data. For similar data from different sources in the data set (such as voltage values ​​measured by multiple sensors), the data is fused using the weighted average method. The weight allocation is based on the credibility of the data source. For example, the weight of a high-precision voltage sensor is set to 0.7, while the weight of an ordinary sensor is set to 0.3. During the fusion process, the weighted average of each group of similar data is calculated separately to eliminate the deviation caused by sensor differences or measurement errors, ensuring the uniformity and accuracy of the data. Finally, the first power sampling data with high consistency is generated.

[0025] Furthermore, in the platform provided in the application embodiment, the forward normalization processing module 11 is also used for: Mining power indicators of power load distribution, wherein the power indicators are marked with weight values; traversing the power indicators, performing indicator type attribution and indicator attribute forward processing, and determining extremely large power indicators, wherein the indicator types include extremely large, extremely small and interval types; matrixing and integrating the extremely large power indicators and performing standardization processing to determine the second power indicator data.

[0026] In an embodiment of the present application, the power indicators of power load distribution are first mined. Core features such as voltage, current, power factor, line loss rate and load rate are collected from the power grid operation data. The relationship between each feature and the load distribution efficiency is calculated using a correlation analysis method (such as the Pearson correlation coefficient). For example, the correlation coefficient of the power factor is 0.8, the correlation coefficient of the load rate is 0.7, and the correlation coefficient of the ambient temperature is only 0.2. The threshold is set to 0.5, and the features with correlations higher than this threshold are used as core indicators of power load distribution. Next, a weight value is assigned to each core indicator to reflect its importance in load distribution, for example, the weight of the power factor is 0.4, and the weight of the line loss rate is 0.3.

[0027] Next, the mined power indicators are traversed to classify the indicator types. Each indicator is classified by threshold rules, and the power factor and power generation capacity are classified as extremely large indicators (the larger the value, the better), the line loss rate and equipment failure rate are classified as extremely small indicators (the smaller the value, the better), and the load rate is classified as an interval indicator (the value is optimal in the range of 70%-90%). Subsequently, the indicator attributes are forward processed. For different types of indicators, corresponding forward conversion methods are used. For extremely small indicators, such as line loss rate, the reciprocal method is used to convert the line loss rate of 0.02 into a positive value of 50; for interval indicators, such as load rate, the interval mapping method is used to map the values ​​within the optimal range (70%-90%) to high scores, for example, 80% is mapped to 0.9, and 50% is mapped to 0.3; for extremely large indicators, their values ​​are kept unchanged because they already have positive characteristics. Through the above processing, all indicators are converted into extremely large power indicators with unified optimization directions, ensuring that different types of indicators have consistent analysis benchmarks.

[0028] Then, the extremely large power indicators are matrix-integrated and standardized, and all the extremely large power indicators that have been forward-processed are integrated into a matrix, in which each column represents an indicator and each row represents a time point or equipment status. Then, the maximum and minimum value standardization method is applied to the integrated matrix to normalize the values ​​of all indicators to the same numerical range, ensuring that different indicators have a unified dimension.

[0029] Through the above steps, the whole process from mining and forward processing of power indicators to matrix integration and standardized processing is completed to generate the second power indicator data.

[0030] The condition initialization module 12 is used to introduce source-load uncertainty conditions, traverse the local power grid topology, perform condition initialization based on the power equipment end's production capacity specificity and the load end's energy consumption specificity, and determine the target uncertainty conditions.

[0031] In an embodiment of the present application, first, the conditional initialization module introduces the source-load uncertainty condition through real-time data extraction and simple statistical analysis. The capacity specificity of the source end is directly obtained from the real-time operation data of the power generation equipment, including the current power generation power and its historical maximum and minimum values. The range of power generation fluctuations is calculated and defined as the capacity uncertainty condition of the source end. At the same time, the energy demand of the load end is extracted from the real-time power consumption record of the load equipment, and the load fluctuation range is obtained by calculating the mean and maximum value of the power consumption data to form the energy uncertainty condition of the load end. Finally, the fluctuation conditions of the source and load ends are integrated to form a preliminary source-load uncertainty condition.

[0032] Next, the topological structure of the local power grid is traversed by graph traversal. The nodes (such as power stations, load centers) and lines (such as cables, transformers) of the local power grid are represented as a graph model, with nodes as vertices and lines as edges. Each node is traversed one by one, the power generation, load and transmission capacity data of the line related to the node are extracted, and the difference between the power generation capacity of the node and the load demand is calculated. At the same time, the transmission capacity and actual load data of each line are recorded, and the key nodes and lines that may be affected by source-load fluctuations are identified.

[0033] After completing the topology analysis, the conditions are initialized directly based on the capacity specificity and energy consumption specificity. The initial condition mapping results are formed by mapping the capacity fluctuation range of the power generation equipment to the power generation node, and mapping the demand fluctuation range of the load equipment to the load node. Combined with the transmission constraints between nodes in the local power grid topology, the uncertainty conditions at the source and load ends are directly associated to generate the initialization conditions.

[0034] Finally, the key target uncertainty conditions are screened out. By checking the condition initialization results, nodes with large power generation fluctuations or drastic changes in load demand, as well as lines with transmission capacity close to the upper limit, are directly marked, and the relevant characteristics of these nodes and lines are used as target uncertainty conditions.

[0035] The model training module 13 is used to supervise the training of the load distribution model by combining the local power grid topology and the target uncertainty condition, wherein the local power grid topology is marked with power line losses, and the load distribution is optimized by lightweight processing and rasterization processing based on expansion and causality.

[0036] In an embodiment of the present application, the model training module supervises the training of the load distribution model by combining the local power grid topology with the target uncertainty conditions. The local power grid topology identifies the power line loss characteristics as the core constraint condition for model training. The power line loss is preset, for example, a loss of 0.003% per meter, that is, during the power grid transmission process, each meter of transmission distance will cause a power loss of 0.003%. This characteristic is embedded in the model as a constraint parameter to correct the transmission loss effect in the distribution strategy. Combined with the load distribution sample data (including power sample data, uncertain condition samples and power distribution samples), a large distribution model is constructed and optimized. The lightweight processing technology is used to determine the lightweight transmission probability and reduce the complexity of the model; at the same time, based on the expansion grid and causal grid division technology, the local characteristics and causal relationships of the power grid operation are captured to optimize the load distribution strategy of the model. Finally, an optimized load distribution model is output.

[0037] Further, such as Figure 2 As shown, in the platform provided by the application embodiment, the model training module 13 is also used for: Call the load distribution sample, and supervise the training of the distribution model with the local power grid topology, the load distribution sample includes power sample data-uncertain condition sample-power distribution sample; based on the preset loss degree, the distribution model is lightweight processed, divided into expansion grid and causal grid, and supervised training of the load distribution model.

[0038] In an embodiment of the present application, first, the load distribution sample is called and combined with the local power grid topology to build a large distribution model. The load distribution sample includes power sample data (such as voltage, current, power factor, etc.), uncertain condition samples (such as the source end power generation fluctuation range and the load end demand change range) and power distribution samples (such as load distribution results under historical conditions). The local power grid topology provides core information such as node distribution, power transmission path and power line loss, which is used as a constraint for model training. For example, the topology clarifies the connection relationship and path loss rate (such as 0.003% per meter) between each node in the power grid. After these data are organized as input features and training labels, a large distribution model is constructed using a supervised learning framework (such as random forest).

[0039] Then, the large allocation model is lightweighted based on the preset loss. By analyzing the logical transfer paths in the model, the contribution of each path to the output result is calculated, and the path contribution is quantified using feature importance analysis methods (such as SHAP value). The pruning threshold is set according to the preset loss, only high-contribution paths are retained, and low-contribution paths are deleted, thereby reducing the complexity of the model while maintaining the model's ability to express core features.

[0040] Subsequently, the expansion grid and the causal grid are divided to optimize the model training. For the expansion grid, based on the time series and spatial distribution characteristics, a clustering algorithm (such as K-means) is used to cluster nodes with similar operating characteristics and divide them into expansion grids. During the training process, the nodes in the expansion grid share the same characteristic weights, which facilitates the optimization of local characteristics. For the causal grid, by constructing a causal relationship network, the causal relationship of power flow is analyzed, and causal analysis methods (such as Granger causality analysis) are used to classify nodes with direct interactions into the causal grid. The causal grid emphasizes the causal relationship between nodes and can enhance the model's ability to express the logic of power transmission.

[0041] Finally, supervised training is performed on the large distribution model. Combined with the division results of lightweight transfer probability, expansion grid and causal grid, the load distribution sample data is input for supervised learning. By continuously optimizing the loss function (such as mean square error) and adjusting the model parameters, the model gradually learns the mapping relationship between the load distribution sample data and the actual distribution results. Through this process, the training of the load distribution model is completed.

[0042] Furthermore, in the platform provided in the application embodiment, the model training module 13 is also used for: Based on the preset loss degree, the lightweight transfer probability is determined, wherein the lightweight transfer probability is the logical transfer probability of the elements of the large distribution model; based on the preset scale characteristics under the time series and spatial distribution, the expansion grid is divided, wherein each grid causality has the same expansion standard; based on the power causal relationship, the causal grid is divided, wherein each causal grid has the same power causal relationship; combined with the lightweight transfer probability, the expansion grid and the causal grid, the large distribution model is supervised and lightweight supervised training is performed to determine the load distribution model.

[0043] In an embodiment of the present application, the lightweight transfer probability is first determined based on a preset loss degree. The preset loss degree defines the range of performance accuracy loss allowed during the model simplification process, and guides the lightweight processing of the model. The contribution of each logical path within the large model to the output result is calculated and allocated through a feature importance analysis method (such as the SHAP value) to obtain a contribution score for the path. Then, the logical paths are sorted from high to low according to the contribution, and a pruning threshold is set according to the preset loss degree. Paths with contributions higher than the threshold are retained, and paths below the threshold are sparsely processed. The lightweight transfer probability is calculated by dividing the contribution by the maximum contribution, which represents the retention probability of each path, thereby streamlining the logical path and reducing the complexity of the model while maintaining the ability to express key features.

[0044] Then, the expansion grid is divided based on the preset scale characteristics of time series and spatial distribution. By analyzing the time series operating characteristics (such as load fluctuations and power generation changes) and spatial distribution characteristics (such as geographical location and connection relationships) of power grid nodes, scale characteristics such as time window length (such as 1 hour) and spatial coverage are set. Clustering algorithms (such as K-means) are used to cluster nodes with similar operating characteristics and divide them into the same expansion grid. In the same expansion grid, nodes share unified feature weights and optimization targets, thereby reducing redundant calculations between nodes within the grid, capturing the local correlation characteristics of the power grid, and improving model training efficiency.

[0045] The causal grid is then divided based on the power causal relationship. By constructing a causal network, the power transmission path and causal relationship between nodes in the power grid are analyzed. Causal analysis methods (such as Granger causality analysis) are used to determine the direct causal relationship between nodes, for example, whether the fluctuation of a certain power generation node significantly affects the change of downstream load nodes. Nodes with causal relationships are divided into the same causal grid to form a causal chain within the grid. Within the causal grid, nodes share the same causal characteristics (such as power flow direction and optimization objectives), which strengthens the model's ability to model power transmission logic, thereby more accurately expressing causal paths.

[0046] Finally, the distribution model is supervised and lightweighted by combining lightweight transfer probability, expansion grid and causal grid. The load distribution samples (including power sample data, uncertain condition samples and distribution samples) are input into the model with the above lightweight transfer probability, expansion grid and causal grid characteristics, and the model parameters are optimized by supervised learning algorithm. In each round of training, the lightweight transfer probability guides the model to simplify the path structure, the expansion grid is used to optimize the expression of local characteristics, and the causal grid is used to enhance the logic modeling of power transmission. By repeatedly iteratively optimizing the loss function (such as mean square error), the model gradually converges and finally generates an accurate and efficient load distribution model.

[0047] Furthermore, in the platform provided in the application embodiment, the model training module 13 is also used for: A first load distribution sample is obtained, transmitted to the distribution large model, and a first soft sample is determined, wherein the first soft sample includes a distribution logic chain; the distribution logic chain is traversed, link node identification and grid marking based on an expansion grid and a causal grid are performed, and a first marked soft sample is determined; based on the lightweight transfer probability, a lightweight training is performed in combination with the first load distribution sample and the first soft sample.

[0048] In an embodiment of the present application, in order to optimize the distribution of the large model and generate the final load distribution model, samples are obtained one by one from the load distribution sample set, and the model performance is gradually optimized by generating a first soft sample and a first labeled soft sample combined with lightweight training, and finally an accurate and efficient load distribution model is generated.

[0049] Specifically, first obtain the first load distribution sample, which contains power sample data (such as node voltage, current, power factor and other basic data describing the operation characteristics of the power grid), uncertain condition samples (such as the fluctuation range of the power generation end and the change range of the load end demand), and power distribution samples (such as the load distribution results under historical conditions). Input these data into the large distribution model, and generate the first soft sample through the initial mapping of the model. The first soft sample contains the distribution logic chain, which describes the logical path of the input features in the model and their transmission relationship, that is, how the input features are transmitted layer by layer through the model to generate the target distribution results.

[0050] Then, the distribution logic chain in the first soft sample is traversed, and the link nodes are rasterized and marked to generate the first marked soft sample. For the expansion grid mark, the time series and spatial characteristics of the logic chain nodes are extracted, and the nodes with similar operating characteristics are grouped into the same expansion grid using a clustering algorithm (such as K-means), and a uniform characteristic weight is assigned to them. This process optimizes the local characteristic expression of the nodes inside the grid and reduces redundant calculations. For the causal grid mark, a causal relationship network is constructed, and the causal path of the nodes in the logic chain is identified through a causal analysis method (such as Granger causality analysis), and the nodes with direct causal relationships are classified into the same causal grid, and a uniform causal path characteristic is assigned. Through the above marking, the first marked soft sample is generated, which integrates the information of the logic chain and the grid characteristics.

[0051] Then, a lightweight training is performed on the large allocation model in combination with the lightweight transfer probability. The lightweight transfer probability is calculated by the path contribution, which indicates the importance of each path in the logic chain to the model output. The probability of path retention is determined by the contribution minus the maximum contribution. High-contribution paths are retained first, and low-contribution paths are sparse or removed. The first load distribution sample and its corresponding first labeled soft sample are input into the model, and the model parameters are optimized through the supervised learning algorithm. In the training, the expansion grid is used to optimize the expression of local characteristics, the causal grid enhances the causal path modeling, and the lightweight processing simplifies the model structure.

[0052] Then repeat the above steps for each sample in the load distribution sample set, generate soft samples from the sample set, label soft samples, and perform lightweight training. Through multiple rounds of training, the large distribution model gradually learns the mapping relationship between input features and target outputs, and the model's logical path and performance expression are significantly optimized.

[0053] Finally, an accurate and efficient load distribution model is generated by training all load distribution samples.

[0054] The power load distribution management module 14 is used to transmit the power sensing data to the load distribution model, make decisions to determine the power load distribution strategy, and perform power load distribution management on the local power grid.

[0055] In an embodiment of the present application, first, the power load distribution strategy is determined based on the load distribution model. The power load distribution management module receives the acquired power perception data (including real-time power status data and standardized power index data) and transmits it to the load distribution model. The load distribution model parses and calculates the input data through its internal logical path and optimization parameters. Specifically, the load distribution model analyzes the characteristics of load demand, power generation fluctuations, uncertain conditions, etc. in the power grid, and combines the power grid topology, power line losses and node causality to generate a distribution plan for each power grid node. The determined power load distribution strategy includes a control plan for the output power of the power generation node and a plan to meet the power demand of the load node, aiming to optimize the operating efficiency and stability of the power grid.

[0056] Subsequently, the local power grid is managed for power load distribution. Based on the generated power load distribution strategy, the module issues specific load dispatching instructions to the power generation equipment and load equipment in the local power grid. For example, on the power generation side, the instructions may include adjusting the power output of the generator set; on the load side, it may involve giving priority to meeting the power needs of critical loads (such as hospitals or industrial equipment), while reducing or delaying the power supply of non-critical loads (such as commercial billboard lighting) to optimize the overall operation efficiency of the power grid under limited resources.

[0057] Furthermore, in the platform provided in the application embodiment, the power load distribution management module 14 is also used for: The first load distribution condition is multi-energy mutual assistance, wherein the multi-energy distribution priority is used as a constraint; the second load distribution condition is multi-microgrid mutual assistance, wherein the local distribution priority is used as a constraint; based on the first load distribution condition and the second load distribution condition, auxiliary power load distribution decision-making is performed.

[0058] In an embodiment of the present application, the first load distribution condition (multi-energy mutual assistance) is first determined. The supply and demand characteristics of various energy forms (such as electricity, wind energy, natural gas, etc.) are extracted, including power generation capacity, energy storage status, and load demand. For energy conversion efficiency, predetermined efficiency parameters are used, which are set based on historical operating data or system calibration values, such as the conversion efficiency of wind power generation and the thermal efficiency of natural gas power generation. By analyzing the energy conversion efficiency and the current supply and demand status, high-efficiency energy forms are prioritized. For example, when electricity demand increases but wind energy resources are sufficient, wind power generation is prioritized to meet demand; when wind power generation is insufficient, natural gas power generation is called to supplement it. By integrating the synergy of multiple energy forms, a multi-energy mutual assistance condition is generated as one of the constraints for load distribution.

[0059] Next, determine the second load distribution condition (multi-microgrid mutual assistance). Analyze the operating characteristics of the microgrids in the local power grid, including power generation capacity (such as photovoltaic, wind power, etc.), energy storage status, and load distribution. By giving priority to dispatching local resources, ensure that the load demand within the microgrid is met first. For example, when the power generation capacity of a microgrid is sufficient, its load demand is first supplied by local resources; when a microgrid is short of power, mutual assistance is achieved by analyzing the resource surplus of adjacent microgrids, and the collaboration between multiple microgrids is dispatched to balance the supply and demand of electricity. The multi-microgrid mutual assistance condition finally generated is based on the priority of local power distribution, which effectively improves the self-sufficiency and flexibility of the local power grid.

[0060] After determining the first load distribution condition and the second load distribution condition, the two are used as constraints to transmit the power sensing data to the load distribution model. The load distribution model comprehensively analyzes the operation status, energy characteristics and load demand of the power grid according to the input power sensing data and constraints, and generates an optimized power load distribution strategy.

[0061] Finally, the local power grid is managed for power load distribution. According to the load distribution strategy, specific instructions are issued to the power generation equipment and load equipment in the power grid, such as requiring wind turbines to increase power output or starting backup natural gas power generation equipment under low wind speed conditions.

[0062] Furthermore, in the platform provided in the application embodiment, the power load distribution management module 14 is also used for: The power load distribution strategy is tracked in response to determine the power distribution response data; the power distribution response data is traversed to verify and trace the abnormal power distribution, and the abnormal power distribution cause is determined, wherein the deviation based on the power distribution amount and the deviation based on the power distribution trend are used as verification standards; based on the abnormal power distribution cause, the power distribution feedback management of the local power grid is performed.

[0063] In the embodiment of the present application, the power load distribution strategy is firstly tracked to determine the distribution response data. The actual power distribution, load response, voltage and current fluctuations, and power output of power generation equipment are collected in real time through sensors and smart meters installed at each node of the power grid. The collected actual operation data is compared with the target data in the distribution strategy node by node, the difference value is calculated, and the distribution response data is generated.

[0064] Then, the distribution response data is traversed to verify and trace the abnormal distribution and determine the cause of the abnormal distribution. Specifically, first, based on the deviation check of the distribution volume, by comparing the actual distribution volume with the target distribution volume, it is determined whether the difference exceeds the preset deviation range, such as ±10%. If the difference of a certain node exceeds this range, it is determined to be an abnormal node. In addition, based on the deviation check of the distribution trend, the time series of the distribution response data is analyzed to determine whether the actual distribution trend is consistent with the load demand change trend. For example, when the load demand increases, the actual distribution volume fails to increase synchronously, and it is marked as a trend abnormality.

[0065] After the abnormal node is identified, the source tracing analysis is performed to check the operating status and data of the relevant upstream and downstream nodes by tracing back the power distribution path. For example, if a node is underpowered, its power generation source and transmission path are gradually traced back to check whether there are problems such as insufficient output of power generation equipment, excessive loss of distribution lines, or overload at the load end, so as to confirm the specific cause of abnormal power distribution.

[0066] Finally, based on the causes of abnormal power distribution, the distribution feedback management of the local power grid is carried out. For short-term anomalies, such as sudden load fluctuations, the distribution strategy is adjusted immediately to quickly replenish power by dispatching backup power generation equipment or adjusting distribution priorities. For long-term anomalies, such as excessive line losses or aging equipment, feedback suggestions are generated, such as optimizing distribution paths, replacing equipment, or upgrading grid infrastructure, and these improvement measures are incorporated into subsequent distribution strategy adjustments. Through closed-loop feedback management, the distribution plan is continuously optimized and the grid operation efficiency is improved.

[0067] In the embodiments of the present application, in summary, the embodiments of the present application have at least the following technical effects: The present application connects a multi-source data interface to determine power perception data, wherein the power perception data includes first power sampling data determined by calling multi-source data related to source and load and performing downsampling fusion, and second power index data determined by performing forward normalization processing on the power index; introducing source-load uncertainty conditions, traversing the local power grid topology, performing conditional initialization based on the production capacity specificity of the power equipment end and the energy consumption specificity based on the load end, and determining the target uncertainty conditions; combining the local power grid topology with the target uncertainty conditions, supervising the training of the load distribution model, wherein the local power grid topology is marked with power line loss, and optimizing the load distribution training by lightweight processing and rasterization processing based on expansion and causality; transmitting the power perception data to the load distribution model, making decisions to determine the power load distribution strategy, and performing power load distribution management on the local power grid. The present invention solves the technical problem that the uncertainty factors of power equipment and load ends are not taken into consideration during the power load distribution process, resulting in poor load distribution optimization effect and unbalanced resource utilization. Through real-time power load distribution optimization based on multi-source sensing data, it can adapt to different load changes and power demands, thereby achieving the technical effect of improving power load distribution efficiency, reducing power line losses and enhancing the operating stability of the local power grid.

[0068] It should be noted that the above-mentioned sequence of the embodiments of the present application is only for description and does not represent the advantages and disadvantages of the embodiments. And the above-mentioned specific embodiments of this specification are described. The processes depicted in the accompanying drawings do not necessarily require the specific order and continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0069] The above description is only a preferred embodiment of the present application and is not intended to limit the present application. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present application shall be included in the protection scope of the present application.

[0070] This specification and drawings are merely exemplary illustrations of the present application and are deemed to cover any and all modifications, variations, combinations or equivalents within the scope of the present application. Obviously, a person skilled in the art may make various modifications and variations to the present application without departing from the scope of the present application. Thus, if these modifications and variations of the present application fall within the scope of the present application and its equivalents, the present application intends to include these modifications and variations.

Claims

1. A multi-source sensing power load distribution platform, characterized in that: The platform includes: A forward normalization processing module, used to connect to a multi-source data interface and determine power sensing data, wherein the power sensing data includes first power sampling data determined by calling multi-source data related to source and load and performing downsampling fusion, and second power index data determined by performing forward normalization processing on the power index; The conditional initialization module is used to introduce source-load uncertainty conditions, traverse the local power grid topology, perform conditional initialization based on the power equipment end's capacity specificity and the load end's energy consumption specificity, and determine the target uncertainty conditions; A model training module is used to supervise the training of a load distribution model by combining a local power grid topology with the target uncertainty condition, wherein the local power grid topology is marked with power line loss, and load distribution optimization training is performed by lightweight processing and grid processing based on expansion and causality; The power load distribution management module is used to transmit the power sensing data to the load distribution model, make decisions to determine the power load distribution strategy, and perform power load distribution management on the local power grid.

2. A multi-source sensing power load distribution platform as claimed in claim 1, characterized in that: The model training module is used to: Calling load distribution samples to supervise and train a large distribution model with the local power grid topology, wherein the load distribution samples include power sample data-uncertain condition samples-power distribution samples; Based on the preset loss degree, the large distribution model is lightweighted, the expansion grid and the causal grid are divided, and the load distribution model is supervised and trained.

3. A multi-source sensing power load distribution platform as claimed in claim 2, characterized in that: The model training module is used to: Based on the preset loss degree, determining a lightweight transfer probability, wherein the lightweight transfer probability is a logical transfer probability of elements of the allocation large model; Based on the preset scale characteristics of time series and spatial distribution, the expansion grid is divided, where each grid has the same expansion standard within the cause and effect; Based on the power causal relationship, the causal grid is divided, wherein each causal grid has the same power causal relationship; The lightweight transfer probability, the expansion grid and the causal grid are combined to perform supervised lightweight supervised training on the large distribution model to determine the load distribution model.

4. A multi-source sensing power load distribution platform as claimed in claim 3, characterized in that: The model training module is used to: Acquire a first load distribution sample, transmit it to the distribution large model, determine a first soft sample, and the first soft sample includes a distribution logic chain; Traversing the distribution logic chain, performing link node identification and grid marking based on the expansion grid and the causal grid, and determining a first marked soft sample; Taking the lightweight transfer probability as a benchmark, a lightweight training is performed in combination with the first load distribution sample and the first soft sample.

5. The multi-source sensing power load distribution platform according to claim 1, characterized in that: The power load distribution management module is used to: The first load distribution condition is to use multi-energy mutual assistance, wherein the multi-energy distribution priority is used as a constraint; The second load distribution condition is to use the mutual assistance of multiple microgrids, wherein the local power distribution priority is used as a constraint; Based on the first load distribution condition and the second load distribution condition, assist in making power load distribution decisions.

6. The multi-source sensing power load distribution platform according to claim 1, characterized in that: The forward normalization processing module is used to: Mining power indicators of power load distribution, wherein the power indicators are marked with weight values; Traversing the power index, performing index type attribution and index attribute forward processing, and determining an extremely large power index, wherein the index type includes extremely large, extremely small, and interval types; The extremely large power index is integrated in a matrix and standardized to determine the second power index data.

7. The multi-source sensing power load distribution platform according to claim 1, characterized in that: The forward normalization processing module is used to: Traversing the multi-source data, performing a first downsampling process on the data granularity based on the data correlation, and determining first downsampling data; Traversing the first down-sampled data, performing a second down-sample process based on the data dimension, and determining the second down-sampled data; The second down-sampled data is subjected to homologous data fusion to determine the first power sampling data.

8. The multi-source sensing power load distribution platform according to claim 1, characterized in that: The power load distribution management module is also used for: Tracking the response of the power load distribution strategy to determine power distribution response data; Traversing the power distribution response data, verifying and tracing the abnormal power distribution, and determining the cause of the abnormal power distribution, wherein the deviation based on the power distribution amount and the deviation based on the power distribution trend are used as verification standards; Based on the abnormal power distribution inducement, power distribution feedback management of the local power grid is performed.

Images