Data intelligent management synchronization system based on power grid material supply chain

Abstract

The invention relates to the technical field of data management, and discloses a power grid material supply chain-based data intelligent management synchronization system, which comprises a weight distribution module, a material demand prediction module, a material allocation scheme preliminary generation module, a data feedback module and a target material synchronization scheme generation module, wherein an attention coefficient matrix of a power grid material supply chain is constructed based on an attenuation weight and a dynamic weight, and weighted fusion is performed on supply data to obtain a material demand prediction result; generating a preliminary material allocation suggestion scheme of the power grid material supply chain; collecting feedback information of the power grid material supply chain after the preliminary material allocation suggestion scheme is executed, and transmitting the feedback information to the preliminary material allocation suggestion scheme; dynamically adjusting the preliminary material allocation suggestion scheme by taking resource constraints and operation priorities of nodes in a power grid material supply chain as constraint conditions to obtain a target material synchronization scheme; the efficiency based on material supply management can be improved.

Description

Technical Field

[0001] This invention relates to the field of data management technology, and in particular to a data-driven intelligent management and synchronization system for power grid material supply chain. Background Technology

[0002] In the field of power grid material supply chain data management, existing technologies have significant limitations in the material demand forecasting stage. They often fail to fully consider the temporal characteristics of historical material consumption data and the dynamic changes in the supply rhythm of supply chain nodes. They mostly use fixed weights to process data, which cannot reduce the impact of invalid or inefficient data due to the distance between historical data and the current time, nor can they adjust data weights according to fluctuations in the supply rhythm of nodes. This results in a lack of accuracy in the integration of supply data and historical consumption data, leading to discrepancies between the material demand forecast results and the actual material demand of the power grid, thus creating potential accuracy risks for subsequent material allocation work.

[0003] Meanwhile, existing technologies are ineffective in generating and optimizing material allocation plans. In the initial plan generation stage, there is insufficient correlation analysis between inventory gaps and the priority of replenishing materials in transit, making it difficult to scientifically differentiate the replenishment demand levels for different gaps. Furthermore, after plan execution, there is a lack of systematic collection and in-depth integration of feedback information, making it impossible to promptly identify constraint conflicts and anomalies during execution. Moreover, the initial plan cannot be dynamically adjusted based on resource constraints and operational priorities of the power grid material supply chain nodes, leading to problems such as resource mismatch and execution obstruction. Ultimately, this reduces the overall management efficiency of the power grid material supply chain and hinders the efficient synchronous scheduling of materials. Summary of the Invention

[0004] This invention provides a data-driven intelligent management and synchronization system for power grid material supply chain to solve the problems mentioned in the background section.

[0005] To achieve the above objectives, the present invention provides a data-driven intelligent management and synchronization system for power grid material supply chain, characterized in that the system includes a weight allocation module, a material demand forecasting module, a preliminary material allocation plan generation module, a data feedback module, and a target material synchronization plan generation module, wherein:

[0006] The weight allocation module is used to assign decay weights to historical material consumption data according to time distance, and to assign dynamic weights to the supply data of the power grid material supply chain according to the rate of change of the supply rhythm of nodes in the power grid material supply chain.

[0007] The material demand forecasting module is used to construct the attention coefficient matrix of the power grid material supply chain based on the attenuation weight and the dynamic weight, and to perform weighted fusion of the supply data based on the attention coefficient matrix to obtain the material demand forecasting result of the power grid material supply chain.

[0008] The preliminary material allocation plan generation module is used to generate a preliminary material allocation suggestion plan for the power grid material supply chain based on the material demand forecast results, the current inventory status data of the power grid material supply chain, and the information on materials in transit.

[0009] The data feedback module is used to collect feedback information from the power grid material supply chain after the execution of the preliminary material allocation proposal, and to transmit the feedback information to the preliminary material allocation proposal.

[0010] The target material synchronization scheme generation module is used to dynamically adjust the preliminary material allocation suggestion scheme based on the resource constraints and operational priorities of the nodes in the power grid material supply chain, so as to obtain the target material synchronization scheme of the power grid material supply chain.

[0011] In a preferred embodiment, when the weight allocation module assigns attenuation weights to historical material consumption data based on time distance and assigns dynamic weights to the supply data of the power grid material supply chain according to the rate of change of the supply rhythm of nodes in the power grid material supply chain, it is specifically used for:

[0012] Obtain the time series of the historical material consumption data and determine the time distance parameter between the historical data point and the current time point;

[0013] Based on the time distance parameter, attenuation weight coefficients are assigned to the historical data points;

[0014] Real-time monitoring of supply rhythm changes at nodes in the power grid material supply chain;

[0015] Based on the changes in the supply rhythm, dynamic weighting coefficients are assigned to the supply data of the power grid material supply chain.

[0016] In a preferred embodiment, when the material demand forecasting module constructs the attention coefficient matrix of the power grid material supply chain based on the attenuation weight and the dynamic weight, it is specifically used for:

[0017] Using vectorized attenuation weights as row vectors and vectorized dynamic weights as column vectors, a matrix outer product is performed to obtain the initial attention coefficient matrix of the power grid material supply chain;

[0018] The initial attention coefficient matrix is ​​normalized to ensure that the element values ​​in the initial attention coefficient matrix are within a standard value range.

[0019] The normalized initial attention coefficient matrix is ​​used as the attention coefficient matrix of the power grid material supply chain.

[0020] In a preferred embodiment, when the material demand forecasting module performs weighted fusion of the supply data based on the attention coefficient matrix to obtain the material demand forecasting result of the power grid material supply chain, it is specifically used for:

[0021] Extract the element values ​​from the attention coefficient matrix as weighting coefficients for the supply data in the power grid material supply chain;

[0022] The weighting coefficients are applied to the corresponding historical consumption data points and real-time supply data points in the power grid material supply chain to obtain the weighted historical consumption dataset and the weighted real-time supply dataset.

[0023] The weighted historical consumption dataset and the weighted real-time supply dataset are fused at the feature level to obtain a multi-dimensional feature vector of the power grid material supply chain;

[0024] The multidimensional feature vector is forward-propagated to obtain the material demand prediction results of the power grid material supply chain.

[0025] In a preferred embodiment, the formula for calculating the material demand forecast result is as follows:

[0026] ;

[0027] In the formula, The forecast results for the aforementioned material demand. For activation function, The weight matrix is ​​obtained by training based on the historical material consumption data. For the multidimensional feature vector, This is the bias vector.

[0028] In a preferred embodiment, when the preliminary material allocation plan generation module generates a preliminary material allocation suggestion plan for the power grid material supply chain based on the material demand forecast results, the current inventory status data of the power grid material supply chain, and the information on materials in transit, it is specifically used for:

[0029] By combining the material demand forecast results with the current inventory status data in the power grid material supply chain, an inventory gap analysis is performed to obtain the out-of-stock nodes and material categories in the power grid material supply chain that have potential out-of-stock risks.

[0030] Based on the in-transit material information of the power grid material supply chain, the estimated arrival time of the materials is estimated to obtain the in-transit material replenishment view of the power grid material supply chain.

[0031] Based on the shortage nodes, the material categories, and the replenishment priorities of the in-transit material replenishment view, an immediate replenishment suggestion is generated for high replenishment priority gaps, and a material transfer suggestion between nodes is generated for low replenishment priority gaps.

[0032] Integrating the replenishment recommendations and the allocation recommendations, a preliminary material allocation plan for the power grid material supply chain is proposed.

[0033] In a preferred embodiment, the preliminary material allocation plan generation module, when performing replenishment priority based on the shortage node, the material category, and the in-transit material replenishment view, generates immediate replenishment suggestions for high replenishment priority gaps and generates inter-node material transfer suggestions for low replenishment priority gaps, is specifically used for:

[0034] For gaps identified as high supply priority, an immediate replenishment suggestion is generated based on the list of qualified suppliers matched with the material category.

[0035] For gaps identified as low supply priority, an internal allocation assessment is initiated. Based on the real-time inventory levels and geographical distribution of nodes in the power grid material supply chain, material allocation recommendations for the low supply priority gaps are generated.

[0036] In a preferred embodiment, when the data feedback module collects feedback information from the power grid material supply chain after implementing the preliminary material allocation proposal and transmits the feedback information to the preliminary material allocation proposal, it is specifically used for:

[0037] Monitor the execution status of nodes in the power grid material supply chain on the preliminary material allocation proposal, and obtain the structured feedback dataset of the nodes;

[0038] Based on the feedback content characteristics of the nodes, the confirmed execution feedback, partial execution feedback, rejected execution feedback, and execution exception feedback in the structured feedback dataset are identified;

[0039] Extract the constraint conflict information and abnormality cause description from the rejection feedback and the execution abnormality feedback to obtain the abnormality feedback detail report of the node;

[0040] The confirmation execution feedback, partial execution feedback, and abnormal feedback details report are integrated to form the power grid material supply chain feedback information;

[0041] The feedback information packet is transmitted in real time to the preliminary generation module of the material allocation plan via a data bus.

[0042] In a preferred embodiment, when the target material synchronization scheme generation module dynamically adjusts the preliminary material allocation suggestion scheme based on resource constraints and operational priorities of nodes in the power grid material supply chain to obtain the target material synchronization scheme for the power grid material supply chain, it is specifically used for:

[0043] A multi-dimensional constraint evaluation system for the power grid material supply chain is established based on the resource constraint parameters and operational priority indicators in the nodes.

[0044] Based on the multi-dimensional constraint evaluation system, resource conflict schemes were identified in the preliminary material allocation proposal.

[0045] Based on the operational priority index, the resource conflict schemes are reordered to obtain the priority ranking list of the power grid material supply chain;

[0046] Based on the priority ranking list, resource allocation among the nodes is coordinated to generate alternative adjustment schemes for the power grid material supply chain;

[0047] A feasibility assessment is performed on the proposed adjustment schemes to determine the optimal adjustment scheme for the resource conflict-prone option.

[0048] The optimal adjustment scheme is integrated with the conflict-free scheme to obtain the target material synchronization scheme of the power grid material supply chain.

[0049] In a preferred embodiment, the calculation formula for the feasibility assessment is as follows:

[0050] ;

[0051] In the formula, The feasibility score of the proposed adjustment scheme is given. To implement cost coefficients, This is the time efficiency coefficient. This is the resource utilization rate coefficient. The estimated total implementation cost of the proposed adjustment scheme is as follows: The minimum estimated implementation cost among the proposed adjustment schemes. The maximum estimated implementation cost among the proposed adjustment options is selected. The minimum estimated completion time among the candidate adjustment schemes. The estimated total completion time for the proposed adjustment scheme is as follows: The maximum estimated completion time among the candidate adjustment schemes is [the highest value]. The comprehensive resource utilization rate of the proposed adjustment scheme is... The minimum overall resource utilization rate among the proposed adjustment schemes. The maximum value of the comprehensive resource utilization rate among the proposed adjustment schemes.

[0052] Compared with the prior art, the present invention has the following beneficial effects:

[0053] 1. This invention assigns attenuation weights to historical material consumption data based on time distance and dynamic weights to supply data based on the rate of change of node supply rhythm using a weight allocation module. A material demand prediction module then constructs an attention coefficient matrix based on these two weights, performing weighted fusion and feature-level fusion on the supply data. Combined with forward propagation calculations, the material demand prediction result is obtained. This technology accurately captures the time value of historical data and the dynamic characteristics of node supply, improving the accuracy of the fusion of supply data and historical consumption data, thereby enhancing the reliability of material demand prediction results and providing accurate data support for subsequent material allocation in the power grid supply chain.

[0054] 2. This invention utilizes a preliminary material allocation plan generation module that combines material demand forecasts, current inventory status, and information on materials in transit to conduct inventory gap analysis, prioritize replenishment, and generate targeted replenishment and allocation suggestions. Simultaneously, a data feedback module collects execution feedback, and a target material synchronous plan generation module dynamically adjusts the preliminary plan based on node resource constraints and operational priorities. This process enables precise control over the entire material allocation plan process, from initial generation to dynamic optimization, effectively improving the scientific validity and feasibility of the allocation plan, ensuring efficient material flow across all nodes of the power grid material supply chain, and ultimately significantly improving the overall efficiency of the power grid material supply chain based on material supply management, achieving synchronous material scheduling. Attached Figure Description

[0055] Figure 1 This is a system architecture diagram of a power grid material supply chain data intelligent management and synchronization system provided in an embodiment of the present invention;

[0056] The realization of the objective, functional features and advantages of the present invention will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0057] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments belong to some, but not all, embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0058] The terminology used in the embodiments of this invention is for the purpose of describing particular embodiments only and is not intended to limit the invention. The singular forms “said” and “the” as used in the embodiments of this invention and the appended claims are also intended to include the plural forms, and “multiple” generally includes at least two unless the context clearly indicates otherwise.

[0059] Depending on the context, the word "if" or "if" as used here can be interpreted as "when," "when," "in response to determination," or "in response to detection." Similarly, depending on the context, the phrase "if determination" or "if detection (of the stated condition or event)" can be interpreted as "when determination," "in response to determination," "when detection (of the stated condition or event)," or "in response to detection (of the stated condition or event)."

[0060] Furthermore, the timing of the steps in the following method embodiments is merely an example and not a strict limitation.

[0061] In practice, the server-side equipment deployed in the power grid material supply chain data intelligent management and synchronization system may consist of one or more devices. This system can be implemented as a business instance, a virtual machine, or hardware devices. For example, it can be implemented as a business instance deployed on one or more devices in a cloud node. Simply put, it can be understood as software deployed on a cloud node to provide the system to various user terminals. Alternatively, it can be implemented as a virtual machine deployed on one or more devices in a cloud node, with application software installed to manage various user terminals. Or, it can also be implemented as a server composed of numerous identical or different types of hardware devices, with one or more hardware devices configured to provide the system to various user terminals.

[0062] In terms of implementation, the intelligent management and synchronization system for power grid material supply chain data and the user terminal are mutually compatible. That is, if the intelligent management and synchronization system for power grid material supply chain data is implemented as an application installed on a cloud service platform, then the user terminal is a client that establishes a communication connection with the application; or if the intelligent management and synchronization system for power grid material supply chain data is implemented as a website, then the user terminal is implemented as a webpage; or if the intelligent management and synchronization system for power grid material supply chain data is implemented as a cloud service platform, then the user terminal is implemented as a mini-program in an instant messaging application.

[0063] like Figure 1The figure shown is a system architecture diagram of a power grid material supply chain data intelligent management and synchronization system provided in an embodiment of the present invention.

[0064] The intelligent management and synchronization system 100 based on power grid material supply chain data described in this invention can be set up in a cloud server. In terms of implementation, it can be used as one or more service devices, or as an application installed in the cloud (e.g., a mobile service operator's server, server cluster, etc.), or it can be developed as a website. Depending on the functions implemented, the intelligent management and synchronization system 100 based on power grid material supply chain data may include a weight allocation module 101, a material demand prediction module 102, a preliminary material allocation plan generation module 103, a data feedback module 104, and a target material synchronization plan generation module 105. The module described in this invention can also be called a unit, which refers to a series of computer program segments that can be executed by an electronic device's processor and can perform a fixed function, stored in the electronic device's memory.

[0065] In this embodiment of the invention, within the intelligent management and synchronization system for power grid material supply chain data, each of the aforementioned modules can be implemented independently and can call upon other modules. This "calling" can be understood as a module connecting to multiple modules of another type and providing corresponding services to those connected modules. In the intelligent management and synchronization system for power grid material supply chain data provided by this embodiment of the invention, the applicability of the system architecture can be adjusted by adding modules and directly calling them without modifying the program code, achieving cluster-based horizontal expansion to quickly and flexibly expand the system. In practical applications, the aforementioned modules can be set up on the same or different devices, or they can be set up on virtual devices, such as service instances on a cloud server.

[0066] The following describes the various components and specific workflows of the intelligent management and synchronization system for power grid material supply chain data, using specific embodiments as examples:

[0067] The weight allocation module 101 is used to assign decay weights to historical material consumption data according to time distance, and to assign dynamic weights to the supply data of the power grid material supply chain according to the rate of change of the supply rhythm of nodes in the power grid material supply chain.

[0068] In this embodiment of the invention, when the weight allocation module assigns attenuation weights to historical material consumption data based on time distance and assigns dynamic weights to the supply data of the power grid material supply chain according to the rate of change of the supply rhythm of nodes in the power grid material supply chain, it is specifically used for:

[0069] Obtain the time series of the historical material consumption data and determine the time distance parameter between the historical data point and the current time point;

[0070] Based on the time distance parameter, attenuation weight coefficients are assigned to the historical data points;

[0071] Real-time monitoring of supply rhythm changes at nodes in the power grid material supply chain;

[0072] Based on the changes in the supply rhythm, dynamic weighting coefficients are assigned to the supply data of the power grid material supply chain.

[0073] Specifically, all material consumption records are retrieved from the historical database of the power grid material supply chain. Each record must include the specific occurrence time and quantity of each consumption data point. These records are arranged in chronological order of occurrence to form a time series of historical material consumption data. After determining the current time point, the time interval between the occurrence time of each historical data point and the current time point is calculated one by one. This time interval is the time distance parameter between the historical data point and the current time point.

[0074] Furthermore, an allocation rule is set where a smaller time distance parameter results in a larger attenuation weight coefficient, and a larger time distance parameter results in a smaller attenuation weight coefficient. For each historical data point, its corresponding time distance parameter is found, and a unique attenuation weight coefficient is matched for that historical data point according to the set rule, thus completing the allocation of the attenuation weight coefficient for the historical data point.

[0075] Furthermore, all nodes in the power grid material supply chain, including production, warehousing, transportation, and distribution, are clearly defined. Real-time data collection points are set up for each node to continuously collect supply-related data such as material output rate, inbound and outbound frequency, transportation cycle, and delivery frequency. The current supply-related data of each node is compared with the supply-related data under recent stable conditions. If a difference occurs, it is determined that the supply rhythm of that node has changed, thus realizing real-time monitoring of changes in the supply rhythm of nodes.

[0076] Furthermore, the range of supply data such as output, inventory, transportation volume, and delivery volume for each node in the power grid material supply chain is determined. Rules are set such that the dynamic weight coefficient of the corresponding supply data increases when the supply rhythm of a node changes positively, decreases when it changes negatively, and maintains the original weight when there is no change. For each piece of supply data, the supply rhythm change of its node is confirmed, and the corresponding dynamic weight coefficient is matched to the supply data according to the rules to complete the allocation of the dynamic weight coefficient of the supply data.

[0077] In summary, assigning attenuation weights to historical material consumption data based on time distance accurately captures the temporal value characteristics of the data. Dynamic attenuation logic strengthens the influence weight of recent, valid data, reduces the interference of outdated data on prediction results, and scientifically releases the temporal relevance of historical consumption data. This weighting method fully leverages the actual reference value of data from different periods, improving the accuracy of historical data applications.

[0078] In summary, assigning dynamic weights to supply data based on the rate of change in node supply rhythm allows for real-time adaptation to the operational fluctuations of supply chain nodes, ensuring that weight adjustments are synchronized with the actual trends in node supply rhythm. This mechanism dynamically responds to the amplitude of fluctuations in node supply rhythm, enhancing the adaptability of supply data to actual operational scenarios. The synergistic effect of these two weighting mechanisms effectively improves the accuracy of integrating supply data with historical consumption data, providing a more reliable data foundation for material demand forecasting and ensuring the effectiveness of power grid material supply chain data applications.

[0079] The material demand forecasting module 102 is used to construct the attention coefficient matrix of the power grid material supply chain based on the attenuation weight and the dynamic weight, and to perform weighted fusion of the supply data based on the attention coefficient matrix to obtain the material demand forecasting result of the power grid material supply chain.

[0080] In this embodiment of the invention, when the material demand forecasting module constructs the attention coefficient matrix of the power grid material supply chain based on the attenuation weight and the dynamic weight, it is specifically used for:

[0081] Using vectorized attenuation weights as row vectors and vectorized dynamic weights as column vectors, a matrix outer product is performed to obtain the initial attention coefficient matrix of the power grid material supply chain;

[0082] The initial attention coefficient matrix is ​​normalized to ensure that the element values ​​in the initial attention coefficient matrix are within a standard value range.

[0083] The normalized initial attention coefficient matrix is ​​used as the attention coefficient matrix of the power grid material supply chain.

[0084] In this embodiment of the invention, when the material demand forecasting module performs weighted fusion of the supply data based on the attention coefficient matrix to obtain the material demand forecasting result of the power grid material supply chain, it is specifically used for:

[0085] Extract the element values ​​from the attention coefficient matrix as weighting coefficients for the supply data in the power grid material supply chain;

[0086] The weighting coefficients are applied to the corresponding historical consumption data points and real-time supply data points in the power grid material supply chain to obtain the weighted historical consumption dataset and the weighted real-time supply dataset.

[0087] The weighted historical consumption dataset and the weighted real-time supply dataset are fused at the feature level to obtain a multi-dimensional feature vector of the power grid material supply chain;

[0088] The multidimensional feature vector is forward-propagated to obtain the material demand prediction results of the power grid material supply chain.

[0089] In this embodiment of the invention, the calculation formula for the material demand forecast result is as follows:

[0090] ;

[0091] In the formula, The forecast results for the aforementioned material demand. For activation function, The weight matrix is ​​obtained by training based on the historical material consumption data. For the multidimensional feature vector, This is the bias vector.

[0092] In summary, by constructing an attention coefficient matrix for the power grid material supply chain using attenuation weights and dynamic weights, an initial matrix is ​​first generated by the vector outer product of the two types of weights. Then, normalization is performed to ensure that the elements are within the standard range. This method can deeply integrate the attenuation characteristics of historical data in the time dimension with the dynamic adaptation characteristics of node supply rhythm, so that each element in the matrix accurately corresponds to the comprehensive weight of specific data, avoiding the limitations of a single weight dimension, and providing a structured and accurate weight basis for data weighting.

[0093] Specifically, first, identify all elements in the vectorized attenuation weight row vector, which correspond to the attenuation weight of each historical data point. Then, identify all elements in the vectorized dynamic weight column vector, which correspond to the dynamic weight of each supply data point. Next, multiply the first element of the row vector with each element of the column vector sequentially to obtain the first product result. Then, multiply the second element of the row vector with each element of the column vector sequentially to obtain the second product result. Repeat this process to multiply all elements of the row vector with all elements of the column vector. Finally, arrange all product results according to the rule that "row vector elements are in the order of rows, and column vector elements are in the order of columns." The resulting matrix is ​​the initial attention coefficient matrix of the power grid material supply chain.

[0094] Furthermore, the standard numerical range is first determined to be 0 to 1. Then, the sum of the values ​​of all elements in the initial attention coefficient matrix is ​​calculated. For each element in the initial attention coefficient matrix, the value of that element is divided by the sum of the values ​​of all elements to obtain a new value for each element. This new value replaces the original values ​​of the elements in the initial attention coefficient matrix, so that all element values ​​in the initial attention coefficient matrix are within the standard numerical range of 0 to 1, thus completing the normalization process of the initial attention coefficient matrix.

[0095] Furthermore, the initial attention coefficient matrix after normalization is directly determined as the attention coefficient matrix of the power grid material supply chain without any additional operations. This matrix is ​​the attention coefficient matrix required for subsequent power grid material supply chain-related analysis or decision-making.

[0096] Specifically, first find each element in the initial attention coefficient matrix, add the value of the first element to the value of the second element, and then add this sum to the value of the third element. Continue to add the new sum to the value of the next element, and repeat this operation until the values ​​of all elements in the matrix are added. The final result is the sum of all elements in the initial attention coefficient matrix.

[0097] Furthermore, we first clarify the specific category of power grid material supply chain data corresponding to each element in the attention coefficient matrix, then directly extract the value of each element in the matrix, and determine the weight coefficient of each extracted element value as the corresponding category of supply data, ensuring that each element value corresponds one-to-one with the corresponding supply data, thus forming a set of weight coefficients for supply data.

[0098] Furthermore, for each historical consumption data point, the corresponding weight coefficient is found, and the value of the historical consumption data point is multiplied by the corresponding weight coefficient to obtain the weighted value of the historical consumption data point. The weighted values ​​of all historical consumption data points are collected and organized to form a weighted historical consumption dataset. For each real-time supply data point, the corresponding weight coefficient is found, and the value of the real-time supply data point is multiplied by the corresponding weight coefficient to obtain the weighted value of the real-time supply data point. The weighted values ​​of all real-time supply data points are collected and organized to form a weighted real-time supply dataset.

[0099] Furthermore, we first sort out the features contained in the weighted historical consumption dataset, such as the weighted consumption values ​​of each time period and the weighted consumption values ​​of different material types. Then we sort out the features contained in the weighted real-time supply dataset, such as the weighted supply values ​​of each node and the weighted supply values ​​of different supply links. Following a fixed order of "historical consumption features first, real-time supply features second", we arrange the features of the two datasets in sequence, so that each feature occupies an independent dimension in the arrangement, forming a vector containing all features. This vector is the multi-dimensional feature vector of the power grid material supply chain.

[0100] Furthermore, each feature value in the multidimensional feature vector is sequentially input into a preset first-level processing stage. The first-level processing stage performs a fixed numerical adjustment on each input feature value, and the adjusted value is passed as the output of the first level to the second-level processing stage. The second-level processing stage performs a fixed numerical adjustment on each received value again and continues to pass the adjusted result to the next level of processing stage. This process is repeated layer by layer until the last level of processing stage. The single value output by the last level of processing stage is the material demand forecast result of the power grid material supply chain.

[0101] Specifically, The weight matrix is ​​trained based on the historical material consumption data. During training, multiple sets of historical material consumption data are first collected. For each set, a corresponding historical multidimensional feature vector is generated using the same method of "feature-level fusion of the weighted historical consumption dataset and the weighted real-time supply dataset." Simultaneously, the actual material demand result corresponding to each set of historical material consumption data is obtained. An initial weight matrix is ​​set, and this initial matrix is ​​used to perform calculations with each set of historical multidimensional feature vectors. The result, plus the initial bias vector, is input into an activation function to obtain the predicted value. The predicted value is compared with the actual material demand result. If a difference exists, the values ​​of the elements in the weight matrix are adjusted, and the calculation and comparison are performed again. This adjustment operation is repeated until the difference between the predicted value and the actual material demand result stabilizes within a preset minimum range. The weight matrix at this point is the final weight matrix. The source of.

[0102] Furthermore, the significance of this formula lies in transforming the multidimensional feature vector into the predicted material demand through a series of operations, first using data trained on historical material consumption data. right Weighting, let Different characteristics in China reflect different degrees of importance according to historical patterns; then use The weighted results are offset to compensate for possible systematic biases in the weighting operation; finally, through... The adjusted result is then converted into a value that conforms to the predicted range. The final output value reflects the material demand in the power grid material supply chain. .

[0103] In summary, when weighting and fusing supply data based on this matrix, matrix elements can be accurately extracted as data weight coefficients, which are applied to historical consumption and real-time supply data respectively. Then, feature-level fusion is used to form a multi-dimensional feature vector, which, combined with forward propagation, yields the prediction result. This process fully activates the synergistic effect of the two types of weights, maximizing the retention of effective data value, filtering out interfering information, improving the accuracy and feature completeness of data fusion, and ultimately outputting prediction results that better meet the actual needs of the power grid material supply chain. This provides reliable data support for the generation of subsequent material allocation plans and ensures the effectiveness of supply chain data applications.

[0104] The preliminary material allocation plan generation module 103 is used to generate a preliminary material allocation suggestion plan for the power grid material supply chain based on the material demand forecast results, the current inventory status data of the power grid material supply chain, and the information on materials in transit.

[0105] In this embodiment of the invention, when the preliminary material allocation plan generation module generates a preliminary material allocation suggestion plan for the power grid material supply chain based on the material demand forecast results, the current inventory status data of the power grid material supply chain, and the information on materials in transit, it is specifically used for:

[0106] By combining the material demand forecast results with the current inventory status data in the power grid material supply chain, an inventory gap analysis is performed to obtain the out-of-stock nodes and material categories in the power grid material supply chain that have potential out-of-stock risks.

[0107] Based on the in-transit material information of the power grid material supply chain, the estimated arrival time of the materials is estimated to obtain the in-transit material replenishment view of the power grid material supply chain.

[0108] Based on the shortage nodes, the material categories, and the replenishment priorities of the in-transit material replenishment view, an immediate replenishment suggestion is generated for high replenishment priority gaps, and a material transfer suggestion between nodes is generated for low replenishment priority gaps.

[0109] Integrating the replenishment recommendations and the allocation recommendations, a preliminary material allocation plan for the power grid material supply chain is proposed.

[0110] In this embodiment of the invention, when the preliminary generation module for the material allocation plan executes the replenishment priority based on the shortage node, the material category, and the in-transit material replenishment view, and generates immediate replenishment suggestions for high replenishment priority gaps and generates inter-node material transfer suggestions for low replenishment priority gaps, it is specifically used for:

[0111] For gaps identified as high supply priority, an immediate replenishment suggestion is generated based on the list of qualified suppliers matched with the material category.

[0112] For gaps identified as low supply priority, an internal allocation assessment is initiated. Based on the real-time inventory levels and geographical distribution of nodes in the power grid material supply chain, material allocation recommendations for the low supply priority gaps are generated.

[0113] Specifically, first, the material demand forecast results and the current inventory status data of each node and each material category in the power grid material supply chain are obtained. The current inventory status data needs to specify the actual inventory quantity of each material category at each node. Then, the material demand forecast results of each node and each material category are compared with the current inventory quantity of that material category at that node. If the current inventory quantity is less than the material demand forecast results, it is determined that there is a potential shortage risk for that material category at that node. Finally, the names of all nodes and material categories with potential shortage risks are recorded to obtain the shortage nodes and material categories with potential shortage risks in the power grid material supply chain.

[0114] Furthermore, information on all goods in transit in the power grid supply chain is first collected. This information should include the type of goods, destination node, current location, mode of transport, and conventional transport speed for each batch of goods in transit. Then, based on the actual distance between the current location and the destination node of each batch of goods in transit, combined with the conventional transport speed of the mode of transport, the transport time required for the batch of goods to travel from the current location to the destination node is calculated. The current time is then added to the transport time to obtain the estimated arrival time of each batch of goods in transit to the destination node. Finally, the type of goods, destination node, and estimated arrival time of each batch of goods in transit are organized into a structured list, with each row in the list corresponding to the information of a batch of goods in transit, thus obtaining the in-transit supply view of the power grid supply chain.

[0115] Furthermore, a supply priority determination rule is first established. The rule is that shortage nodes that are key nodes for ensuring regional power supply and shortages of materials that are emergency materials affecting the normal operation of the power grid are classified as high supply priority. Shortage nodes that are not key nodes and shortages of materials that are not emergency materials are classified as low supply priority. Then, for high supply priority shortages, the system checks whether there are corresponding materials in the in-transit supply view that are expected to arrive in time to meet the demand for the shortage. If not, an immediate replenishment suggestion is generated, specifying the material category to be replenished, the replenishment quantity, and the target shortage node. For low supply priority shortages, the system checks whether other non-shortage nodes in the power grid supply chain have redundant inventory of the material category. If so, a material transfer suggestion between nodes is generated, specifying the transferring node, the transferring node, the type of material to be transferred, and the transfer quantity.

[0116] Furthermore, all generated immediate replenishment suggestions are first categorized by material type, with each material type summarizing the corresponding replenishment nodes, replenishment quantities, and replenishment requirements. Then, all generated material allocation suggestions are categorized by transfer-out nodes, with each transfer-out node summarizing the corresponding transfer-in nodes, allocated material types, and transfer quantities. Finally, the categorized immediate replenishment suggestions and material allocation suggestions are integrated into a single document, clearly indicating the type of each suggestion, the names of the involved nodes, the name of the material type, the specific quantity, and the key execution points, thus forming the preliminary material allocation suggestion scheme for the power grid material supply chain.

[0117] Specifically, first, identify the specific material category corresponding to the high-priority supply gap. Then, retrieve the pre-associated list of qualified suppliers for that material category. The list of qualified suppliers must include information such as supplier name, maximum quantity available for that material category, historical delivery cycle, and contact information. Next, calculate the replenishment quantity for the high-priority supply gap by subtracting the current inventory quantity at the corresponding shortage node from the material demand forecast for that material category. Then, select the supplier with the shortest historical delivery cycle and a supply quantity no less than the replenishment quantity from the list of qualified suppliers. Finally, generate an immediate replenishment suggestion for the selected supplier. The suggestion must specify the material category, replenishment quantity, target shortage node, selected supplier name, and the time limit for initiating the replenishment request.

[0118] Furthermore, the real-time inventory levels of all nodes in the power grid material supply chain for the material categories identified as low-priority supply gaps are first collected. The real-time inventory levels must include the actual inventory quantity of each node for that material category and the node's own material demand forecast. The inventory redundancy of each node for that material category is calculated, and nodes with an inventory redundancy greater than 0 are selected as potential transfer nodes. Then, the geographical distribution information of these potential transfer nodes and the shortage nodes corresponding to the low-priority supply gaps is obtained. The transportation distance from each potential transfer node to the shortage node is calculated, and the potential transfer node with the shortest transportation distance is selected as the final transfer node. The transfer quantity is determined, and finally, a material transfer suggestion is generated. The suggestion must clearly specify the material category, transfer quantity, name of the transfer node, name of the transfer receiving node, and the suggested transfer transportation method.

[0119] In summary, generating preliminary material allocation recommendations based on material demand forecasts, current inventory status data, and information on materials in transit allows for the collaborative application of multi-dimensional data, avoiding the biased allocation resulting from relying on a single data source. By combining forecast results with current inventory data to conduct inventory gap analysis, it is possible to accurately identify nodes and material categories with potential stockout risks, clarify the core objectives of allocation needs, provide precise guidance for subsequent allocation actions, and prevent resource waste caused by ambiguous demand positioning.

[0120] In summary, by estimating arrival times based on information on materials in transit and generating a replenishment view, the dynamics of resources awaiting warehousing can be monitored in real time. This provides a more comprehensive resource reference for gap replenishment, avoiding duplicate replenishment or omissions of resources in transit. Furthermore, by generating targeted suggestions based on replenishment priorities, high-priority gaps can be replenished immediately, while low-priority gaps can be planned through internal allocation. This enables differentiated and efficient resource distribution, ensuring that critical gaps are met first. The final integrated preliminary plan has a clear allocation direction and specific measures, laying a reliable foundation for subsequent plan optimization and execution, and ensuring the orderly commencement of power grid material supply chain allocation.

[0121] The data feedback module 104 is used to collect feedback information from the power grid material supply chain after the execution of the preliminary material allocation suggestion plan, and transmit the feedback information to the preliminary material allocation suggestion plan.

[0122] In this embodiment of the invention, when the data feedback module collects feedback information from the power grid material supply chain after implementing the preliminary material allocation proposal and transmits the feedback information to the preliminary material allocation proposal, it is specifically used for:

[0123] Monitor the execution status of nodes in the power grid material supply chain on the preliminary material allocation proposal, and obtain the structured feedback dataset of the nodes;

[0124] Based on the feedback content characteristics of the nodes, the confirmed execution feedback, partial execution feedback, rejected execution feedback, and execution exception feedback in the structured feedback dataset are identified;

[0125] Extract the constraint conflict information and abnormality cause description from the rejection feedback and the execution abnormality feedback to obtain the abnormality feedback detail report of the node;

[0126] The confirmation execution feedback, partial execution feedback, and abnormal feedback details report are integrated to form the power grid material supply chain feedback information;

[0127] The feedback information packet is transmitted in real time to the preliminary generation module of the material allocation plan via a data bus.

[0128] Specifically, a fixed execution status feedback period is set for each node in the power grid material supply chain. Each node is required to submit the execution status information of the preliminary material allocation proposal within the period. The feedback information must include the node name, the type of the corresponding proposal (immediate replenishment or material allocation), the quantity executed, the quantity not executed, and a description of the execution progress. All the information submitted by all nodes is organized according to the dimension of "node name - proposal type". Each record contains a complete execution status field, resulting in a structured feedback dataset for the node.

[0129] Furthermore, a set of criteria for judging the characteristics of feedback content is established. The characteristics of confirmed execution feedback are that the feedback content explicitly includes statements such as "completely executed as suggested" or "all execution processes have been initiated" without mentioning any unexecuted parts; the characteristics of partial execution feedback are that the feedback content includes statements such as "number of parts executed" or "only some steps completed" and specifies the number or steps that were not executed; the characteristics of rejection feedback are that the feedback content includes statements such as "cannot execute the suggestion" or "does not meet the conditions for execution" with a reason for rejection; the characteristics of execution anomaly feedback are that the feedback content includes statements such as "problems occurred during execution" or "execution was interrupted" and describes the abnormal situation. Each piece of feedback content in the structured feedback dataset is examined against these criteria to identify the corresponding confirmed execution feedback, partial execution feedback, rejection feedback, and execution anomaly feedback.

[0130] Furthermore, each refusal to execute feedback and execution anomaly feedback is examined one by one. From the refusal to execute feedback, content involving conflicts of resource, time, and other constraints, such as "insufficient inventory to meet allocation quantity" and "supplier unable to deliver on time," is extracted as constraint conflict information. From the execution anomaly feedback, descriptions of reasons for non-execution or execution anomalies, such as "transportation route obstructed" and "personnel shortage," are extracted as anomaly cause descriptions. From the execution anomaly feedback, constraint conflict information, such as "materials damaged in transit resulting in insufficient quantity" and "node receiving process failure," is extracted. Descriptions of anomalies, such as "natural disasters affecting transportation" and "equipment failure delaying warehousing," are extracted. The node name, feedback type, constraint conflict information, and anomaly cause description corresponding to each feedback are compiled into a record in a unified format. All records are summarized to form an anomaly feedback detail report for the node.

[0131] Furthermore, the confirmed execution feedback is categorized by node, with each node recording the corresponding suggestion type, the specific content of the confirmed execution, and the execution completion time (if completed); partial execution feedback is categorized by node, with each node recording the corresponding suggestion type, the number of executed items, the number of unexecuted items, and the reason for unexecution; each record in the abnormal feedback detail report is integrated into the feedback content of that node, ensuring that various types of feedback information from the same node are presented centrally, forming complete power grid material supply chain feedback information.

[0132] Furthermore, in accordance with the data packet format required by the preliminary generation module of the material allocation plan, the power grid material supply chain feedback information is encapsulated into a feedback information packet. This data packet includes the generation time of the feedback information, the number of nodes involved, the statistical quantity of various types of feedback, and detailed feedback records. The data bus transmission channel is activated to establish a connection with the preliminary generation module of the material allocation plan. The transmission trigger condition is set to the completion of the feedback information packet encapsulation. After the trigger, the data bus sends the feedback information packet to the preliminary generation module of the material allocation plan in real time to ensure that the module receives the feedback information packet in a timely manner.

[0133] In summary, collecting feedback information after the implementation of preliminary material allocation proposals allows for real-time monitoring of the implementation status at each node of the power grid material supply chain. This generates a structured feedback dataset and accurately identifies different feedback types, such as confirmed implementation, partial implementation, rejection, and implementation anomalies. This comprehensive approach captures the real-world situation during implementation, preventing misjudgments of execution effectiveness due to missing information. Furthermore, by extracting constraint conflict information and anomaly descriptions from rejection and anomaly feedback, detailed anomaly feedback reports can be generated. This allows for in-depth analysis of the root causes of implementation obstacles, providing specific problem-solving guidance for subsequent optimization, rather than merely collecting superficial execution results.

[0134] In summary, transmitting the integrated feedback information to the initial material allocation proposal creates a data loop of "proposal execution - feedback collection - information feedback," providing clear data support for adjustments to the initial proposal and avoiding blind optimization. This mechanism allows subsequent dynamic adjustments to the proposal to precisely match the actual execution conditions of nodes, ensuring that the adjusted proposal better aligns with the operational realities of the power grid material supply chain. This lays a reliable foundation for the generation of synchronous proposals for target materials, guaranteeing the smooth progress of the supply chain allocation process and the accuracy of management decisions.

[0135] The target material synchronization scheme generation module 105 is used to dynamically adjust the preliminary material allocation suggestion scheme based on the resource constraints and operational priorities of the nodes in the power grid material supply chain, so as to obtain the target material synchronization scheme of the power grid material supply chain.

[0136] In this embodiment of the invention, when the target material synchronization scheme generation module dynamically adjusts the preliminary material allocation suggestion scheme based on the resource constraints and operational priorities of the nodes in the power grid material supply chain to obtain the target material synchronization scheme of the power grid material supply chain, it is specifically used for:

[0137] A multi-dimensional constraint evaluation system for the power grid material supply chain is established based on the resource constraint parameters and operational priority indicators in the nodes.

[0138] Based on the multi-dimensional constraint evaluation system, resource conflict schemes were identified in the preliminary material allocation proposal.

[0139] Based on the operational priority index, the resource conflict schemes are reordered to obtain the priority ranking list of the power grid material supply chain;

[0140] Based on the priority ranking list, resource allocation among the nodes is coordinated to generate alternative adjustment schemes for the power grid material supply chain;

[0141] A feasibility assessment is performed on the proposed adjustment schemes to determine the optimal adjustment scheme for the resource conflict-prone option.

[0142] The optimal adjustment scheme is integrated with the conflict-free scheme to obtain the target material synchronization scheme of the power grid material supply chain.

[0143] In this embodiment of the invention, the calculation formula for the feasibility assessment is as follows:

[0144] ;

[0145] In the formula, The feasibility score of the proposed adjustment scheme is given. To implement cost coefficients, This is the time efficiency coefficient. This is the resource utilization rate coefficient. The estimated total implementation cost of the proposed adjustment scheme is as follows: The minimum estimated implementation cost among the proposed adjustment schemes. The maximum estimated implementation cost among the proposed adjustment options is selected. The minimum estimated completion time among the candidate adjustment schemes. The estimated total completion time for the proposed adjustment scheme is as follows: The maximum estimated completion time among the candidate adjustment schemes is [the highest value]. The comprehensive resource utilization rate of the proposed adjustment scheme is... The minimum overall resource utilization rate among the proposed adjustment schemes. The maximum value of the comprehensive resource utilization rate among the proposed adjustment schemes.

[0146] Specifically, the resource constraint parameters in the nodes are first clarified, including resource-limiting indicators such as the node's maximum inventory capacity, the number of available transport vehicles, the maximum supply from suppliers, and the number of warehouse workers. The operational priority indicators specifically include the power supply importance level of the node's guaranteed area (high level for core areas and low level for ordinary areas) and the degree of impact of power grid equipment failures on the corresponding materials (high level for affecting the main grid operation and low level for affecting branches). Then, quantitative evaluation standards are set for the resource constraint parameters. For example, if the maximum inventory capacity is lower than the required quantity of the plan item, it is determined that the constraint is triggered. Level classification standards are set for the operational priority indicators. For example, the indicator level of the power supply node in the core area is higher than that of the ordinary area. Finally, the evaluation standards of the resource constraint parameters and the level standards of the operational priority indicators are integrated to form a multi-dimensional constraint evaluation system for the power grid material supply chain that covers resource constraints and priority determination.

[0147] Furthermore, each item in the preliminary material allocation proposal is extracted one by one. Each item must include the nodes involved, the type of resources required (inventory, transportation, supply, etc.), and the quantity of resources. By comparing the resource constraint parameters in the multi-dimensional constraint evaluation system, it is checked whether the quantity of resources required for each item exceeds the upper limit of the resource constraint parameters of the corresponding node. For example, if an allocation item requires node A to allocate 50 transformers, but the maximum number of transformers that can be allocated in node A's resource constraint parameters is 40, then the item triggers a resource constraint. All items that trigger resource constraints are summarized to obtain the item with resource conflict in the preliminary material allocation proposal.

[0148] Furthermore, the operational priority index level corresponding to each resource conflicting scheme item is obtained, and the sorting rule is set so that scheme items with higher operational priority index levels are ranked first and those with lower levels are ranked last. If there are scheme items with the same level, the emergency use time limit of the materials involved in the scheme items is further compared, and those with shorter time limits are ranked first. All resource conflicting scheme items are arranged in order according to this rule, and the arrangement result is presented in the form of a list. The list includes the scheme item name, the involved nodes, the operational priority level, and the arrangement number, thus obtaining the priority ranking list of the power grid material supply chain.

[0149] Furthermore, based on the priority ranking list, resources are allocated first to the top-ranked resource conflict scheme item. For example, if this scheme item requires storage space from node B, unused storage resources on node B are prioritized to meet its needs. For subsequent scheme items, if the originally planned resources are occupied by higher-priority scheme items, the resource constraint parameters of other nodes are re-queried. For example, the original plan to transfer materials from node C is changed to transfer from node D (node ​​D's inventory of the material has not reached the constraint limit), or the resource usage time is adjusted (the resources are called after the higher-priority scheme item has used up its resources). The adjusted resource allocation method for each resource conflict scheme item is compiled into an independent scheme, and all independent schemes are summarized to form the alternative adjustment schemes for the power grid material supply chain.

[0150] Furthermore, for each alternative adjustment plan, the resource allocation is re-examined against the multi-dimensional constraint evaluation system to confirm whether there are still issues with resource constraint parameters exceeding the upper limit, such as whether the number of transport vehicles required for the adjusted plan is within the range of available vehicles at the corresponding node; at the same time, the degree to which the adjusted plan meets the operational priority indicators is evaluated, such as whether the material needs of high-priority nodes can still be guaranteed after the adjustment; alternative adjustment plans with resource conflicts or those that cannot meet operational priority requirements are excluded, and the plan with the highest resource utilization rate (such as inventory call volume approaching the node redundancy limit) and the least impact on other plans is selected from the remaining plans as the optimal adjustment plan for the plan with resource conflicts.

[0151] Furthermore, firstly, conflict-free scheme items that do not trigger resource constraints in the preliminary material allocation proposal are screened out, and it is confirmed that the resource allocation, involved nodes, and execution requirements of these scheme items all comply with the multi-dimensional constraint evaluation system; then, the original resource conflict scheme items corresponding to the optimal adjustment scheme are replaced with the content of the optimal adjustment scheme, ensuring that the replaced scheme items have no resource conflicts; finally, the conflict-free scheme items and the replaced optimal adjustment scheme are integrated according to the original classification logic of the preliminary material allocation proposal (such as by replenishment and allocation type) to form a complete scheme that includes all scheme items and has no resource conflicts, thus obtaining the target material synchronization scheme of the power grid material supply chain.

[0152] Specifically, To implement the cost coefficient, a fixed value is set based on the cost control target of the power grid material supply chain. If cost control is a priority in operation, a larger value is set, and if the cost control requirement is lower, a smaller value is set. The time efficiency coefficient is a fixed value set based on the timeliness requirements of the power grid material supply chain. If rapid allocation is required, a larger value is set; if the timeliness requirements are less stringent, a smaller value is set. The resource utilization coefficient is a fixed value set based on the resource optimization goals of the power grid material supply chain. If it is necessary to prioritize improving resource utilization, a larger value is set; if the resource utilization requirements are lower, a smaller value is set.

[0153] Furthermore, the significance of this formula lies in quantifying the feasibility of alternative adjustment plans from three dimensions: cost, time, and resource utilization. First, addressing the cost dimension: subtract the minimum estimated implementation cost of all alternatives from the estimated total implementation cost of a single plan, then divide by the difference between the maximum and minimum estimated implementation costs to obtain a standardized cost result. Subtracting this result from 1 increases the score as the cost decreases, and multiplying by an implementation cost coefficient reflects the impact of cost on feasibility. Next, addressing the time dimension: subtract the estimated total completion time of a single plan from the minimum estimated completion time of all alternatives, then divide by the maximum estimated completion time. The difference between the time and the minimum value yields a time-standardized result, where shorter time results in higher scores. This is then multiplied by a time efficiency coefficient to reflect the impact of time on feasibility. Next, the resource utilization dimension is processed: the minimum overall resource utilization rate of all candidates is subtracted from the overall resource utilization rate of a single option, and then divided by the difference between the maximum and minimum overall resource utilization rates to obtain a resource utilization standardized result. Higher utilization rates result in higher scores, and this is then multiplied by a resource utilization rate coefficient to reflect the impact of resource utilization on feasibility. Finally, the results of the three dimensions are summed. The higher the feasibility score, the more feasible the adjusted option is.

[0154] In summary, adjusting the initial allocation plan based on the resource constraints and operational priorities of the power grid material supply chain nodes can accurately identify resource conflicts in the initial plan by establishing a multi-dimensional constraint evaluation system. This avoids obstacles to plan implementation caused by neglecting the actual resource capabilities of nodes, ensuring that the adjustment direction always aligns with the actual boundaries of supply chain operations, and providing precise targets for plan optimization.

[0155] In summary, re-prioritizing conflicting items based on operational priorities ensures that resource allocation prioritizes core operational needs, guaranteeing the timely fulfillment of allocation requirements for high-priority nodes or critical materials, aligning with the operational priorities of the power grid supply chain. Further coordination of resource allocation and feasibility assessment (combining cost, time efficiency, and resource utilization to select the optimal solution) ensures that the adjusted plan is both scientifically sound and practically feasible. The resulting target material synchronization plan achieves precise matching between node resources and allocation needs, ensuring the synchronization of material dispatch and effectively improving the efficiency and accuracy of material management and dispatching in the power grid supply chain.

[0156] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the present invention can be implemented in other specific forms without departing from the spirit or essential characteristics of the present invention.

[0157] The embodiments of this application can acquire and process relevant data based on artificial intelligence technology. Artificial intelligence is the theory, method, technology, and application system that uses digital computers or machines controlled by digital computers to simulate, extend, and expand human intelligence, perceive the environment, acquire knowledge, and use that knowledge to obtain optimal results.

[0158] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention.

Claims

1. A data-driven intelligent management and synchronization system for power grid material supply chain, characterized in that: The system includes a weight allocation module, a material demand forecasting module, a preliminary material allocation plan generation module, a data feedback module, and a target material synchronization plan generation module, wherein: The weight allocation module is used to assign decay weights to historical material consumption data according to time distance, and to assign dynamic weights to the supply data of the power grid material supply chain according to the rate of change of the supply rhythm of nodes in the power grid material supply chain. The material demand forecasting module is used to construct the attention coefficient matrix of the power grid material supply chain based on the attenuation weight and the dynamic weight, and to perform weighted fusion of the supply data based on the attention coefficient matrix to obtain the material demand forecasting result of the power grid material supply chain. The preliminary material allocation plan generation module is used to generate a preliminary material allocation suggestion plan for the power grid material supply chain based on the material demand forecast results, the current inventory status data of the power grid material supply chain, and the information on materials in transit. The data feedback module is used to collect feedback information from the power grid material supply chain after the execution of the preliminary material allocation proposal, and to transmit the feedback information to the preliminary material allocation proposal. The target material synchronization scheme generation module is used to dynamically adjust the preliminary material allocation suggestion scheme based on the resource constraints and operational priorities of the nodes in the power grid material supply chain, so as to obtain the target material synchronization scheme of the power grid material supply chain.

2. The intelligent management and synchronization system for power grid material supply chain data as described in claim 1, characterized in that, When the weight allocation module assigns attenuation weights to historical material consumption data based on time distance and assigns dynamic weights to the supply data of the power grid material supply chain according to the rate of change of the supply rhythm of nodes in the power grid material supply chain, it is specifically used for: Obtain the time series of the historical material consumption data and determine the time distance parameter between the historical data point and the current time point; Based on the time distance parameter, attenuation weight coefficients are assigned to the historical data points; Real-time monitoring of supply rhythm changes at nodes in the power grid material supply chain; Based on the changes in the supply rhythm, dynamic weighting coefficients are assigned to the supply data of the power grid material supply chain.

3. The intelligent management and synchronization system for power grid material supply chain data as described in claim 1, characterized in that, When the material demand forecasting module constructs the attention coefficient matrix of the power grid material supply chain based on the attenuation weight and the dynamic weight, it is specifically used for: Using vectorized attenuation weights as row vectors and vectorized dynamic weights as column vectors, a matrix outer product is performed to obtain the initial attention coefficient matrix of the power grid material supply chain; The initial attention coefficient matrix is ​​normalized to ensure that the element values ​​in the initial attention coefficient matrix are within a standard value range. The normalized initial attention coefficient matrix is ​​used as the attention coefficient matrix of the power grid material supply chain.

4. The intelligent management and synchronization system for power grid material supply chain data as described in claim 3, characterized in that, When the material demand forecasting module performs weighted fusion of the supply data based on the attention coefficient matrix to obtain the material demand forecasting result of the power grid material supply chain, it is specifically used for: Extract the element values ​​from the attention coefficient matrix as weighting coefficients for the supply data in the power grid material supply chain; The weighting coefficients are applied to the corresponding historical consumption data points and real-time supply data points in the power grid material supply chain to obtain the weighted historical consumption dataset and the weighted real-time supply dataset. The weighted historical consumption dataset and the weighted real-time supply dataset are fused at the feature level to obtain a multi-dimensional feature vector of the power grid material supply chain; The multidimensional feature vector is forward-propagated to obtain the material demand prediction results of the power grid material supply chain.

5. The intelligent management and synchronization system for power grid material supply chain data as described in claim 4, characterized in that, The formula for calculating the forecast results of material demand is as follows: ； In the formula, The forecast results for the aforementioned material demand. For activation function, The weight matrix is ​​obtained by training based on the historical material consumption data. For the multidimensional feature vector, This is the bias vector.

6. The intelligent management and synchronization system for power grid material supply chain data as described in claim 1, characterized in that, When the preliminary material allocation plan generation module generates a preliminary material allocation suggestion plan for the power grid material supply chain based on the material demand forecast results, the current inventory status data of the power grid material supply chain, and the information on materials in transit, it is specifically used for: By combining the material demand forecast results with the current inventory status data in the power grid material supply chain, an inventory gap analysis is performed to obtain the out-of-stock nodes and material categories in the power grid material supply chain that have potential out-of-stock risks. Based on the in-transit material information of the power grid material supply chain, the estimated arrival time of the materials is estimated to obtain the in-transit material replenishment view of the power grid material supply chain. Based on the shortage nodes, the material categories, and the replenishment priorities of the in-transit material replenishment view, an immediate replenishment suggestion is generated for high replenishment priority gaps, and a material transfer suggestion between nodes is generated for low replenishment priority gaps. Integrating the replenishment recommendations and the allocation recommendations, a preliminary material allocation plan for the power grid material supply chain is proposed.

7. The intelligent management and synchronization system for power grid material supply chain data as described in claim 6, characterized in that, The preliminary material allocation plan generation module, when executing the replenishment priority based on the shortage node, the material category, and the in-transit material replenishment view, generates immediate replenishment suggestions for high replenishment priority gaps and generates inter-node material transfer suggestions for low replenishment priority gaps, is specifically used for: For gaps identified as high supply priority, an immediate replenishment suggestion is generated based on the list of qualified suppliers matched with the material category. For gaps identified as low supply priority, an internal allocation assessment is initiated. Based on the real-time inventory levels and geographical distribution of nodes in the power grid material supply chain, material allocation recommendations for the low supply priority gaps are generated.

8. The intelligent management and synchronization system for power grid material supply chain data as described in claim 1, characterized in that, When the data feedback module collects feedback information from the power grid material supply chain after implementing the preliminary material allocation proposal, and transmits the feedback information to the preliminary material allocation proposal, it is specifically used for: Monitor the execution status of nodes in the power grid material supply chain on the preliminary material allocation proposal, and obtain the structured feedback dataset of the nodes; Based on the feedback content characteristics of the nodes, the confirmed execution feedback, partial execution feedback, rejected execution feedback, and execution exception feedback in the structured feedback dataset are identified; Extract the constraint conflict information and abnormality cause description from the rejection feedback and the execution abnormality feedback to obtain the abnormality feedback detail report of the node; The confirmation execution feedback, partial execution feedback, and abnormal feedback details report are integrated to form the power grid material supply chain feedback information; The feedback information packet is transmitted in real time to the preliminary generation module of the material allocation plan via a data bus.

9. The intelligent management and synchronization system for power grid material supply chain data as described in claim 8, characterized in that, When the target material synchronization scheme generation module dynamically adjusts the preliminary material allocation suggestion scheme based on resource constraints and operational priorities of nodes in the power grid material supply chain to obtain the target material synchronization scheme for the power grid material supply chain, it is specifically used for: A multi-dimensional constraint evaluation system for the power grid material supply chain is established based on the resource constraint parameters and operational priority indicators in the nodes. Based on the multi-dimensional constraint evaluation system, resource conflict schemes were identified in the preliminary material allocation proposal. Based on the operational priority index, the resource conflict schemes are reordered to obtain the priority ranking list of the power grid material supply chain; Based on the priority ranking list, resource allocation among the nodes is coordinated to generate alternative adjustment schemes for the power grid material supply chain; A feasibility assessment is performed on the proposed adjustment schemes to determine the optimal adjustment scheme for the resource conflict-prone option. The optimal adjustment scheme is integrated with the conflict-free scheme to obtain the target material synchronization scheme of the power grid material supply chain.

10. The intelligent management and synchronization system for power grid material supply chain data as described in claim 9, characterized in that, The calculation formula for the feasibility assessment is as follows: ； In the formula, The feasibility score of the proposed adjustment scheme is given. To implement cost coefficients, This is the time efficiency coefficient. This is the resource utilization rate coefficient. The estimated total implementation cost of the proposed adjustment scheme is as follows: The minimum estimated implementation cost among the proposed adjustment schemes. The maximum estimated implementation cost among the proposed adjustment options is selected. The minimum estimated completion time among the candidate adjustment schemes. The estimated total completion time for the proposed adjustment scheme is as follows: The maximum estimated completion time among the candidate adjustment schemes is [the highest value]. The comprehensive resource utilization rate of the proposed adjustment scheme is... The minimum overall resource utilization rate among the proposed adjustment schemes. The maximum value of the comprehensive resource utilization rate among the proposed adjustment schemes.

Images