Power grid optimization scheduling method and system based on intelligent optimization algorithm

Abstract

The invention discloses a power grid optimization scheduling method and system based on an intelligent optimization algorithm, and relates to the technical field of intelligent power grids, and the method comprises the steps: constructing an intelligent power grid energy model, and determining a total income value set of a user; acquiring a dual variable set, a transaction energy mismatching limit value set and an energy transmission space variable set of the user; performing embedded coding on a total revenue value set, a dual variable set, a transaction energy mismatching limit value set and an energy transmission space variable set of the user by using the embedded matrix; obtaining an economic allocation quota multi-mode collaborative self-interaction representation vector; based on the economic allocation quota multi-modal collaborative self-interaction representation vector, generating an economic allocation quota value of the user; and scheduling based on the economic allocation quota value of the user to complete power grid optimization scheduling. The calculation accuracy and rationality of the economic dispatching quota value are effectively improved, and more accurate economic dispatching quota reference is provided for users.

Description

Technical Field

[0001] The present invention relates to the field of smart grid technology, and in particular to a grid optimization scheduling method and system based on an intelligent optimization algorithm. Background Art

[0002] With the public's growing demand for low-carbon energy conservation, the traditional power system is undergoing a transformation towards intelligence. By installing distributed power generation equipment (such as solar panels, small wind turbines, etc.) in homes, users can effectively reduce distribution and transmission losses and make full use of renewable energy. Since renewable energy generation facilities generate excess electricity during the power generation process, in order to promote the development of renewable energy, users are encouraged to sell the remaining renewable energy to others through the existing power grid to obtain benefits.

[0003] The invention patent with the publication number CN113283655A proposes a distributed smart grid energy dispatching method based on a consensus optimization algorithm, which collects information such as the user's real-time indoor temperature, flexible load value, and energy transaction mark, and combines the preset energy supply and marketing balance conditions to build a smart grid energy model. Based on the difference between the user's real-time indoor temperature and the preset ideal temperature, the user's discomfort cost function set is calculated, and the user's total benefit value set is calculated by accumulating the user's discomfort cost function set, the energy transaction price and value set between the user and the power grid system and other users. Then, the proximal alternating direction multiplier consensus algorithm (ADMM) is used, based on the user's total benefit value set, combined with the dual variables in the communication process between the user and the neighboring node, and the transaction energy mismatch amount between users (that is, when the energy transaction between users is carried out, the difference between the actual transaction energy and the expected transaction energy) After a series of iterative calculations, the user's economic dispatch quota value is obtained. However, since the physical transmission distance of electric energy between users will lead to different degrees of energy loss and cost increase, this factor is not considered in the calculation process of the above economic dispatch quota value, which may lead to errors in the calculation of the economic dispatch quota value. Summary of the invention

[0004] The purpose of the present invention is to overcome the defects in the prior art that the physical transmission distance of electric energy between users will lead to different degrees of energy loss and cost increase, resulting in errors in the calculation of the user's economic dispatch quota value, and reducing the user's electricity sales revenue. A power grid optimization dispatching method and system based on an intelligent optimization algorithm are provided, which take into account the influence of the physical transmission distance of electric energy of the user device on the economic dispatch quota value, and improve the accuracy and rationality of the economic dispatch quota value.

[0005] The purpose of the present invention is achieved through the following technical solutions: A power grid optimization dispatching method based on an intelligent optimization algorithm comprises: constructing a smart grid energy model, and determining a user's total benefit value set according to the smart grid energy model; and obtaining a user's dual variable set, a transaction energy mismatch quota value set, and an energy transmission space variable set; Use the embedding matrix to embed the user's total benefit value set, dual variable set, transaction energy mismatch value set and energy transmission space variable set to obtain the total benefit embedding coding vector, dual variable embedding coding vector, transaction energy mismatch value embedding coding vector and energy transmission space embedding coding vector; Perform cross-domain query interaction coding on the total revenue embedding coding vector, dual variable embedding coding vector, transaction energy mismatch quota embedding coding vector and energy transmission space embedding coding vector to obtain the multimodal collaborative self-interaction representation vector of economic allocation quota; Generate the user's economic allocation quota value based on the multimodal collaborative self-interaction representation vector of the economic allocation quota; Dispatching is carried out based on the user's economic allocation quota value to achieve optimized grid dispatching.

[0006] In this scheme, the dual variable set is used to characterize the communication variables during point-to-point communication between the users, the transaction energy mismatch amount value set is used to characterize the numerical value of energy mismatch when users conduct energy transactions with each other, and the energy transmission space variable set is used to characterize the electric energy transmission distance value between users.

[0007] This solution introduces the information of power transmission distance between users, and uses artificial intelligence technology based on deep learning to perform semantic embedding encoding on the user's total revenue set, dual variable set, transaction energy mismatch quota value set and energy transmission space variable set, so as to obtain the semantic embedding feature representation of each information, and then use the user's total revenue information as the main reference basis, conduct cross-domain query matching analysis with other relevant factors, and mine the collaborative interaction characteristics of multimodal information to decode and calculate the user's economic allocation quota value. Therefore, it can effectively improve the calculation accuracy and rationality of the economic dispatch quota value, and provide users with a more accurate reference for the economic allocation quota.

[0008] Preferably, the total revenue embedding coding vector, the dual variable embedding coding vector, the transaction energy mismatching quota embedding coding vector and the energy transmission space embedding coding vector are cross-domain queried interactively to obtain the economic allocation quota multimodal collaborative self-interactive representation vector, including: The total revenue embedding coding vector is used as the query feature vector, and the feature matrix composed of the total revenue embedding coding vector, the dual variable embedding coding vector, the transaction energy mismatch quota embedding coding vector and the energy transmission space embedding coding vector is used as the key matrix. The query feature vector and the key matrix are input into the cross-domain feature gating query network with local correlation enhanced by the key matrix to obtain the multimodal collaborative self-interaction representation vector of the economic allocation quota.

[0009] Preferably, the total revenue embedding coding vector is used as the query feature vector, the feature matrix composed of the total revenue embedding coding vector, the dual variable embedding coding vector, the transaction energy mismatch quota embedding coding vector and the energy transmission space embedding coding vector is used as the key matrix, and the query feature vector and the key matrix are input into the cross-domain feature gating query network with local correlation enhancement of the key matrix to obtain the multimodal collaborative self-interaction representation vector of the economic allocation quota, including: Performing linear transformation on each row vector in the key matrix to obtain an initial key matrix; Calculating feature distribution difference energy coefficients between the query feature vector and each row vector in the initial key matrix to obtain a set of feature distribution difference energy coefficients; Comparing each feature distribution difference energy coefficient in the set of feature distribution difference energy coefficients with a predetermined threshold, and in response to the feature distribution difference energy coefficient being less than the predetermined threshold, taking a mean vector between a row vector corresponding to the feature distribution difference energy coefficient and the query feature vector as an updated row vector to obtain an updated key matrix; Processing the query feature vector using the value embedding matrix to obtain a value vector; Using the query feature vector as the query vector and each row vector in the update key matrix as the key vector, the query vector, the key vector and the value vector are input into a cross-domain attention query module based on a Transformer structure to obtain a sequence of cross-domain query attention vectors of economic allocation quotas; The positional mean vector of the sequence of the cross-domain query focus vector of the economic allocation quota is calculated to obtain a multimodal collaborative self-interaction representation vector of the economic allocation quota.

[0010] Preferably, the step of calculating the characteristic distribution difference energy coefficients between the query characteristic vector and each row vector in the initial key matrix to obtain a set of characteristic distribution difference energy coefficients includes: Extracting semantic differences between the query feature vector and each row vector in the initial key matrix respectively to obtain a set of semantically related difference feature vectors; The energy coefficient of each semantic association difference feature vector in the set of semantic association difference feature vectors is calculated to obtain a set of feature distribution difference energy coefficients.

[0011] Preferably, the step of respectively extracting the semantic differences between the query feature vector and each row vector in the initial key matrix to obtain a set of semantically associated difference feature vectors includes: Calculate the position point addition, position point subtraction and position point multiplication between the query feature vector and the row vector to obtain a first matching fusion result, a second matching fusion result and a third matching fusion result; After the first matching fusion result, the second matching fusion result and the third matching fusion result are cascaded into a multi-scale semantic association matching vector, one-dimensional convolution processing and maximum pooling processing are performed on the multi-scale semantic association matching vector to obtain a semantic association difference feature vector.

[0012] Preferably, the step of calculating the energy coefficient of each semantic association difference feature vector in the set of semantic association difference feature vectors to obtain a set of feature distribution difference energy coefficients includes: The square of the difference between the maximum eigenvalue and the eigenmean of the semantic association difference feature vector, the feature variance of the semantic association difference feature vector, and the bias term are weightedly summed to obtain a weighted sum value, and the sum of the feature variance and the bias term of the semantic association difference feature vector is divided by the weighted sum value to obtain the feature distribution difference energy coefficient.

[0013] Preferably, the method of generating the economic allocation quota value of the user based on the multimodal collaborative self-interaction representation vector of the economic allocation quota includes: The multimodal collaborative self-interaction representation vector of the economic allocation quota is input into the decoder to obtain the decoded value of the user's economic allocation quota value.

[0014] Preferably, the step of inputting the multimodal collaborative self-interaction representation vector of the economic allocation quota into a decoder to obtain a decoded value of the economic allocation quota value of the user comprises: The multimodal collaborative self-interaction representation vector of the economic allocation quota is input into the decoder for decoding calculation, and the multi-layer neural network structure of the decoder model is used to perform layer-by-layer feature extraction and nonlinear mapping on the multimodal collaborative self-interaction representation vector of the economic allocation quota to capture the multi-source data association information contained therein, and generate a decoded value of the economic allocation quota value of the user according to the user's needs and constraints in the economic dispatch of the power grid.

[0015] A power grid optimization dispatching system based on an intelligent optimization algorithm, based on a power grid optimization dispatching method based on an intelligent optimization algorithm, comprising: A smart grid energy model building module is used to build a smart grid energy model; A total benefit value set determination module, used to determine the total benefit value set of each user according to the smart grid energy model; The economic allocation quota value determination module is used to determine the economic allocation quota value of each user based on the total income value set of each user.

[0016] Preferably, the economic allocation quota value determination module includes: A data collection unit, used to obtain a user's total revenue set, a dual variable set, a transaction energy mismatch value set, and an energy transmission space variable set; An embedding coding unit, used for embedding the user's total benefit set, dual variable set, transaction energy mismatch amount value set and energy transmission space variable set using an embedding matrix to obtain a total benefit embedding coding vector, a dual variable embedding coding vector, a transaction energy mismatch amount embedding coding vector and an energy transmission space embedding coding vector; A cross-domain query interaction coding unit, used for performing cross-domain query interaction coding on the total revenue embedding coding vector, the dual variable embedding coding vector, the transaction energy mismatch quota embedding coding vector and the energy transmission space embedding coding vector to obtain a multi-modal collaborative self-interaction representation vector of the economic allocation quota; The user's economic allocation quota value determination unit is used to generate the user's economic allocation quota value based on the economic allocation quota multimodal collaborative self-interaction representation vector.

[0017] The beneficial effect of the present invention is that the present invention introduces the power transmission distance information between users, and uses artificial intelligence technology based on deep learning to perform semantic embedding coding on the user's total revenue set, dual variable set, transaction energy mismatch quota value set and energy transmission space variable set, so as to obtain the semantic embedding feature representation of each information, and then use the user's total revenue information as the main reference basis, perform cross-domain query matching analysis with other relevant factors, and mine the collaborative interaction characteristics of multimodal information, so as to perform decoding calculation of the user's economic allocation quota value. Therefore, the calculation accuracy and rationality of the economic dispatch quota value can be effectively improved, and a more accurate reference for the economic allocation quota can be provided to users. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] Figure 1 is a flow chart of the present invention; Figure 2 It is a system block diagram of the present invention.

[0019] Among them: 201, smart grid energy model construction module, 202, total benefit value set determination module, 203, economic allocation quota value determination module, 2031, data collection unit, 2032, embedded coding unit, 2033, cross-domain query interaction coding unit, 2034, economic allocation quota value determination unit. DETAILED DESCRIPTION

[0020] Example embodiments will now be described more fully with reference to the accompanying drawings. However, example embodiments can be implemented in a variety of forms and should not be construed as limited to the examples set forth herein; rather, these embodiments are provided so that this application will be more comprehensive and complete and fully convey the concept of the example embodiments to those skilled in the art.

[0021] In addition, described feature, structure or characteristic can be combined in one or more embodiments in any suitable manner. In the following description, many specific details are provided to provide a full understanding of the embodiments of the present application. However, those skilled in the art will appreciate that the technical scheme of the present application can be put into practice without one or more of the specific details, or other methods, components, devices, steps, etc. can be adopted. In other cases, known methods, devices, realizations or operations are not shown or described in detail to avoid blurring the various aspects of the application.

[0022] The block diagrams shown in the accompanying drawings are merely functional entities and do not necessarily correspond to physically independent entities. That is, these functional entities may be implemented in software form, or in one or more hardware modules or integrated circuits, or in different networks and / or processor devices and / or microcontroller devices.

[0023] The flowcharts shown in the accompanying drawings are only exemplary and do not necessarily include all the contents and operations / steps, nor must they be executed in the order described. For example, some operations / steps can be decomposed, and some operations / steps can be combined or partially combined, so the actual execution order may change according to actual conditions.

[0024] Example: A power grid optimization dispatching method based on intelligent optimization algorithm, such as Figure 1 As shown, including: Step S101, constructing a smart grid energy model; Step S102, determining a total benefit value set of a user according to the smart grid energy model; Step S103, determining the economic allocation quota value of the user according to the total income value set of the user; According to the user's total income value set, the user's economic allocation quota value is determined, including: Step S1031, obtaining the user's dual variable set, transaction energy mismatch value set and energy transmission space variable set; Step S1032, using the embedding matrix to embed the user's total benefit value set, dual variable set, transaction energy mismatch value set and energy transmission space variable set to obtain a total benefit embedding coding vector, a dual variable embedding coding vector, a transaction energy mismatch value embedding coding vector and an energy transmission space embedding coding vector; Step S1033, performing cross-domain query interactive coding on the total revenue embedding coding vector, the dual variable embedding coding vector, the transaction energy mismatch quota embedding coding vector, and the energy transmission space embedding coding vector to obtain a multi-modal collaborative self-interaction representation vector of the economic allocation quota; Step S1034, generating the economic allocation quota value of the user based on the multimodal collaborative self-interaction representation vector of the economic allocation quota; Dispatching is carried out based on the user's economic allocation quota value to achieve optimized grid dispatching.

[0025] In this scheme, the dual variable set is used to characterize the communication variables during point-to-point communication between the users, the transaction energy mismatch amount value set is used to characterize the numerical value of energy mismatch when users conduct energy transactions with each other, and the energy transmission space variable set is used to characterize the electric energy transmission distance value between users.

[0026] This solution introduces the information of power transmission distance between users, and uses artificial intelligence technology based on deep learning to perform semantic embedding encoding on the user's total revenue set, dual variable set, transaction energy mismatch quota value set and energy transmission space variable set, so as to obtain the semantic embedding feature representation of each information, and then use the user's total revenue information as the main reference basis, conduct cross-domain query matching analysis with other relevant factors, and mine the collaborative interaction characteristics of multimodal information to decode and calculate the user's economic allocation quota value. Therefore, it can effectively improve the calculation accuracy and rationality of the economic dispatch quota value, and provide users with a more accurate reference for the economic allocation quota.

[0027] The cross-domain query interactive coding of the total revenue embedding coding vector, the dual variable embedding coding vector, the transaction energy mismatch quota embedding coding vector and the energy transmission space embedding coding vector is performed to obtain the multimodal collaborative self-interaction representation vector of the economic allocation quota, including: The total revenue embedding coding vector is used as the query feature vector, and the feature matrix composed of the total revenue embedding coding vector, the dual variable embedding coding vector, the transaction energy mismatch quota embedding coding vector and the energy transmission space embedding coding vector is used as the key matrix. The query feature vector and the key matrix are input into the cross-domain feature gating query network with local correlation enhanced by the key matrix to obtain the multimodal collaborative self-interaction representation vector of the economic allocation quota.

[0028] The method uses the total revenue embedding coding vector as the query feature vector, uses the feature matrix composed of the total revenue embedding coding vector, the dual variable embedding coding vector, the transaction energy mismatch quota embedding coding vector and the energy transmission space embedding coding vector as the key matrix, inputs the query feature vector and the key matrix into the cross-domain feature gating query network with local correlation enhancement of the key matrix to obtain the multi-modal collaborative self-interaction representation vector of the economic allocation quota, including: Performing linear transformation on each row vector in the key matrix to obtain an initial key matrix; Calculating feature distribution difference energy coefficients between the query feature vector and each row vector in the initial key matrix to obtain a set of feature distribution difference energy coefficients; Comparing each feature distribution difference energy coefficient in the set of feature distribution difference energy coefficients with a predetermined threshold, and in response to the feature distribution difference energy coefficient being less than the predetermined threshold, taking a mean vector between a row vector corresponding to the feature distribution difference energy coefficient and the query feature vector as an updated row vector to obtain an updated key matrix; Processing the query feature vector using the value embedding matrix to obtain a value vector; Using the query feature vector as the query vector and each row vector in the update key matrix as the key vector, the query vector, the key vector and the value vector are input into a cross-domain attention query module based on a Transformer structure to obtain a sequence of cross-domain query attention vectors of economic allocation quotas; The positional mean vector of the sequence of the cross-domain query focus vector of the economic allocation quota is calculated to obtain a multimodal collaborative self-interaction representation vector of the economic allocation quota.

[0029] The step of calculating the characteristic distribution difference energy coefficients between the query characteristic vector and each row vector in the initial key matrix to obtain a set of characteristic distribution difference energy coefficients includes: Extracting semantic differences between the query feature vector and each row vector in the initial key matrix respectively to obtain a set of semantically related difference feature vectors; The energy coefficient of each semantic association difference feature vector in the set of semantic association difference feature vectors is calculated to obtain a set of feature distribution difference energy coefficients.

[0030] The step of respectively extracting the semantic differences between the query feature vector and each row vector in the initial key matrix to obtain a set of semantically associated difference feature vectors includes: Calculate the position point addition, position point subtraction and position point multiplication between the query feature vector and the row vector to obtain a first matching fusion result, a second matching fusion result and a third matching fusion result; After the first matching fusion result, the second matching fusion result and the third matching fusion result are cascaded into a multi-scale semantic association matching vector, one-dimensional convolution processing and maximum pooling processing are performed on the multi-scale semantic association matching vector to obtain a semantic association difference feature vector.

[0031] The step of calculating the energy coefficient of each semantic association difference feature vector in the set of semantic association difference feature vectors to obtain a set of feature distribution difference energy coefficients includes: The square of the difference between the maximum eigenvalue and the eigenmean of the semantic association difference feature vector, the feature variance of the semantic association difference feature vector, and the bias term are weightedly summed to obtain a weighted sum value, and the sum of the feature variance and the bias term of the semantic association difference feature vector is divided by the weighted sum value to obtain the feature distribution difference energy coefficient.

[0032] The method of generating the economic allocation quota value of the user based on the multimodal collaborative self-interaction representation vector of the economic allocation quota includes: The multimodal collaborative self-interaction representation vector of the economic allocation quota is input into the decoder to obtain the decoded value of the user's economic allocation quota value.

[0033] The step of inputting the multimodal collaborative self-interaction representation vector of the economic allocation quota into a decoder to obtain a decoded value of the economic allocation quota value of the user includes: The multimodal collaborative self-interaction representation vector of the economic allocation quota is input into the decoder for decoding calculation, and the multi-layer neural network structure of the decoder model is used to perform layer-by-layer feature extraction and nonlinear mapping on the multimodal collaborative self-interaction representation vector of the economic allocation quota to capture the multi-source data association information contained therein, and generate a decoded value of the economic allocation quota value of the user according to the user's needs and constraints in the economic dispatch of the power grid.

[0034] In order to comprehensively capture the various factors that affect the economic dispatch quota value of the user, in this embodiment, first, the user's total benefit set, dual variable set, transaction energy mismatch quota value set and energy transmission space variable set are obtained. Among them, the total benefit set reflects the overall economic benefits obtained by the user from the power grid or energy trading, and provides the user's energy consumption habits and economic incentives. The dual variable set is used to characterize the communication variables during the point-to-point communication between the user and other users, reflecting the energy transaction constraints during the communication process between the user and the neighboring node. The transaction energy mismatch quota value set is used to characterize the numerical value of the energy mismatch when the user conducts energy transactions with other users, indicating the gap between the actual transaction volume and the expected transaction volume of the user when conducting energy transactions. The energy transmission space variable set is used to characterize the value of the electric energy transmission distance between the user and other users, reflecting the line loss of electric energy transmission. By obtaining these multi-source information, the economic incentives, communication status, transaction matching and transmission costs of users in the economic dispatch of the power grid can be fully understood, thereby providing a solid data foundation for subsequent optimization scheduling decisions.

[0035] Considering that different data sets may have different data types and dimensions, directly using deep learning algorithms for data joint analysis may lead to inaccurate analysis results due to the dimensional differences between different data sets. Therefore, in the technical solution of the present application, an embedding matrix is ​​further used to embed the user's total revenue set, dual variable set, transaction energy mismatch value set, and energy transfer space variable set, and each data set is mapped to a unified semantic feature space to eliminate the dimensional differences between different data sets and capture the intrinsic semantic information of each data set, thereby obtaining a total revenue embedding coding vector, a dual variable embedding coding vector, a transaction energy mismatch value embedding coding vector, and an energy transfer space embedding coding vector.

[0036] Furthermore, the core goal of optimizing scheduling is to maximize total revenue or minimize total cost, because total revenue is the core driving force for users to participate in the economic scheduling of power grids. Therefore, in this embodiment, the total revenue information of users is used as the main reference basis, and other relevant factors are combined for collaborative interaction analysis to calculate the economic allocation quota. This embodiment introduces a cross-domain feature gated query network based on local association enhancement of key matrix, which uses the total revenue embedding coding vector as the query feature vector, and the feature matrix composed of the total revenue embedding coding vector, the dual variable embedding coding vector, the transaction energy mismatch quota embedding coding vector and the energy transmission space embedding coding vector as the key matrix, and performs query matching analysis on the query feature vector and the key matrix, and considers the influence of each factor on the total revenue to mine the collaborative interaction characteristics of multimodal information. In this way, it can be ensured that when calculating the economic allocation quota value of the user, the influence of each factor on the total revenue can be fully considered, thereby improving the accuracy and rationality of the calculation results.

[0037] Specifically, in calculating the position mean vector of the sequence of the cross-domain query attention vector of the economic allocation quota to obtain the multimodal collaborative self-interaction representation vector of the economic allocation quota, the matching formula of the cross-domain query is: , , , , , , , , , in, represents the bond matrix, , , and are the row vectors in the key matrix, namely, the total revenue embedding coding vector, the dual variable embedding coding vector, the transaction energy mismatch amount embedding coding vector and the energy transfer space embedding coding vector, , represents the first row vectors, is the linear transformation matrix, , and are different bias terms, represents the initial bond matrix, , , and are the row vectors in the initial key matrix, represents the query feature vector, represents dot product, Indicates point addition, Indicates point reduction, Indicates cascade operation, represents the maximum pooling operation, represents a one-dimensional convolution operation, Indicates semantic association difference feature vectors, , and Respectively represent the The feature variance, maximum eigenvalue and feature mean of the semantic association difference feature vectors, is the first in the set of characteristic distribution difference energy coefficients The characteristic distribution difference energy coefficient, is the predetermined threshold, represents the updated key matrix, represents the first row vectors, is the value embedding matrix, is a value vector, is the normalized exponential function, The value of is the length of the row vector in the update key matrix, represents the transpose of a vector, It is The cross-domain query attention vector of economic allocation quota, Represents the multimodal collaborative self-interaction representation vector of the economic allocation quota.

[0038] In this embodiment, the multimodal collaborative self-interaction representation vector of the economic allocation quota is input into the decoder to obtain the decoded value of the user's economic allocation quota value, specifically including: Determine the maximum eigenvalue of the multimodal collaborative self-interaction of the economic allocation quota and the minimum eigenvalue of the multimodal collaborative self-interaction of the economic allocation quota of the multimodal collaborative self-interaction representation vector of the economic allocation quota; Calculate the mean value of the multimodal collaborative self-interaction of the economic allocation quota and the standard deviation of the multimodal collaborative self-interaction of the economic allocation quota of the feature set of the multimodal collaborative self-interaction representation vector of the economic allocation quota, and calculate the quotient of the mean value of the multimodal collaborative self-interaction of the economic allocation quota and the standard deviation of the multimodal collaborative self-interaction of the economic allocation quota to obtain the statistical normalization value of the multimodal collaborative self-interaction of the economic allocation quota; Calculate the inverse of each eigenvalue of the economic allocation quota multimodal collaborative self-interaction representation vector, multiply it by the difference between the maximum eigenvalue of the economic allocation quota multimodal collaborative self-interaction and the minimum eigenvalue of the economic allocation quota multimodal collaborative self-interaction, and then subtract it from the statistical normalization value of the economic allocation quota multimodal collaborative self-interaction to obtain an economic allocation quota multimodal collaborative self-interaction distribution approximation vector; Calculate an exponential function with a natural constant as the base, with each eigenvalue of the economic allocation quota multimodal collaborative self-interactive distribution approximation vector as the exponent to obtain an economic allocation quota multimodal collaborative self-interactive distribution class approximation vector; Performing dot-wise addition of the economic allocation quota multimodal collaborative self-interaction distribution class approximation vector and the economic allocation quota multimodal collaborative self-interaction statistical normalized value, and calculating the base-two logarithm of the absolute value of each eigenvalue of the dot-wise addition vector to obtain an optimized economic allocation quota multimodal collaborative self-interaction representation vector; and The optimized economic allocation quota multimodal collaborative self-interaction representation vector is input into a decoder to obtain a decoded value of the user's economic allocation quota value.

[0039] The specific formula is as follows: , , , , is the standardized value of the multimodal collaborative self-interaction statistics of the economic allocation quota, and is the difference between the maximum eigenvalue of the multimodal collaborative self-interaction of the economic allocation quota and the minimum eigenvalue of the multimodal collaborative self-interaction of the economic allocation quota, is the logarithm to base 2, is the maximum eigenvalue of the multimodal collaborative self-interaction of the economic allocation quota of the multimodal collaborative self-interaction representation vector of the economic allocation quota, is the minimum eigenvalue of the multimodal collaborative self-interaction of the economic allocation quota of the multimodal collaborative self-interaction representation vector of the economic allocation quota, is the economic allocation quota multimodal collaborative self-interaction mean of the feature set of the economic allocation quota multimodal collaborative self-interaction representation vector, is the standard deviation of the multimodal collaborative self-interaction of the economic allocation quota of the feature set of the multimodal collaborative self-interaction representation vector of the economic allocation quota, is the reciprocal of each eigenvalue of the multimodal collaborative self-interaction representation vector of the economic allocation quota, is the approximate vector of the multimodal collaborative self-interaction distribution of economic allocation quota, It is an exponential function that calculates the exponential of each eigenvalue of a vector with a natural constant as the base. is the optimized multi-modal collaborative self-interaction representation vector of economic allocation quota, It is the point multiplication by position. It is added by location point. It is subtracted by position point.

[0040] In this embodiment, the decoding formula is: ,in, is the optimized multi-modal collaborative self-interaction representation vector of economic allocation quota, It is the decoded value of the user's economic allocation quota. is the weight matrix, Represents matrix multiplication.

[0041] A power grid optimization dispatching system based on an intelligent optimization algorithm, based on a power grid optimization dispatching method based on an intelligent optimization algorithm, such as Figure 2 As shown, including: A smart grid energy model building module 201 is used to build a smart grid energy model; A total benefit value set determination module 202, configured to determine a total benefit value set for each user according to the smart grid energy model; The economic allocation quota value determination module 203 is used to determine the economic allocation quota value of each user according to the total revenue value set of each user.

[0042] The economic allocation quota value determination module includes: The data collection unit 2031 is used to obtain the user's total income set, dual variable set, transaction energy mismatch value set and energy transmission space variable set; The embedding coding unit 2032 is used to embed the user's total benefit set, dual variable set, transaction energy mismatch value set and energy transmission space variable set using an embedding matrix to obtain a total benefit embedding coding vector, a dual variable embedding coding vector, a transaction energy mismatch value embedding coding vector and an energy transmission space embedding coding vector; A cross-domain query interaction coding unit 2033 is used to perform cross-domain query interaction coding on the total revenue embedding coding vector, the dual variable embedding coding vector, the transaction energy mismatch quota embedding coding vector and the energy transmission space embedding coding vector to obtain a multi-modal collaborative self-interaction representation vector of the economic allocation quota; The user's economic allocation quota value determination unit 2034 is used to generate the user's economic allocation quota value based on the economic allocation quota multimodal collaborative self-interaction representation vector.

[0043] In summary, since the total revenue embedding coding vector, the dual variable embedding coding vector, the transaction energy mismatch amount embedding coding vector and the energy transmission space embedding coding vector respectively represent the high-dimensional semantic embedding coding features of the total revenue set, the dual variable set, the transaction energy mismatch amount value set and the energy transmission space variable set, when the total revenue embedding coding vector is used as the query feature vector, the feature matrix composed of the total revenue embedding coding vector, the dual variable embedding coding vector, the transaction energy mismatch amount embedding coding vector and the energy transmission space embedding coding vector is obtained by the embedding coding as the key matrix, and it is input into the cross-domain feature gated query network with enhanced local correlation of the key matrix, the local correlation difference of the key matrix attributed to the difference in the distribution of the encoded semantic features will also cause the driving query difference across the semantic space, so that the economic allocation quota multimodal collaborative self-interaction representation vector will also have a diversified set expression distribution of cross-domain query aggregation features. Therefore, it is expected to improve the balance between the regression mapping accuracy and completeness when the economic allocation quota multimodal collaborative self-interaction representation vector is input into the decoder for decoding regression, thereby improving the accuracy of the decoded value of the user's economic allocation quota value.

[0044] Therefore, by performing random statistical normalization on the diverse feature set of the multimodal collaborative self-interaction representation vector of the economic allocation quota, the confidence space constructed based on the overall eigenvalue of the multimodal collaborative self-interaction representation vector of the economic allocation quota is approximated by a standardized continuous probability density distribution relative to the response hypothesis test of each eigenvalue of the multimodal collaborative self-interaction representation vector of the economic allocation quota, thereby establishing the target reachability of the diverse feature distribution of the multimodal collaborative self-interaction representation vector of the economic allocation quota to a unified regression target, so as to achieve a balanced executable performance between mapping accuracy and mapping completeness in the decoding regression process based on the diverse feature distribution of the multimodal collaborative self-interaction representation vector of the economic allocation quota, and improve the accuracy of the decoded value of the user's economic allocation quota value obtained by the multimodal collaborative self-interaction feature of the economic allocation quota to the input decoder.

[0045] The above description is only a preferred embodiment of the present application and an explanation of the technical principles used. Those skilled in the art should understand that the scope of disclosure involved in the present application is not limited to the technical solution formed by a specific combination of the above technical features, but should also cover other technical solutions formed by any combination of the above technical features or their equivalent features without departing from the above disclosed concept. For example, the above features are replaced with the technical features with similar functions disclosed in this application (but not limited to) by each other to form a technical solution.

[0046] In addition, although each operation is described in a specific order, this should not be construed as requiring these operations to be performed in the specific order shown or to be performed in a sequential order. Under certain circumstances, multitasking and parallel processing may be advantageous. Similarly, although some specific implementation details are included in the above discussion, these should not be construed as limiting the scope of the application. Some features described in the context of a separate embodiment can also be implemented in a single embodiment in combination. On the contrary, the various features described in the context of a single embodiment can also be implemented in multiple embodiments individually or in any suitable sub-combination mode.

[0047] Although the subject matter has been described in language specific to structural features and / or method logic actions, it should be understood that the subject matter defined is not necessarily limited to the specific features or actions described above. On the contrary, the specific features and actions described above are merely exemplary forms of implementation. With respect to the apparatus in the above-mentioned embodiments, the specific manner in which each module performs the operation has been described in detail in the embodiments related to the method, and will not be described in detail here.

Claims

1. A power grid optimization dispatching method based on an intelligent optimization algorithm, comprising: Constructing a smart grid energy model, and determining a user's total benefit value set based on the smart grid energy model; wherein the user's dual variable set, transaction energy mismatch value set, and energy transmission space variable set are also obtained; Use the embedding matrix to embed the user's total benefit value set, dual variable set, transaction energy mismatch value set and energy transmission space variable set to obtain the total benefit embedding coding vector, dual variable embedding coding vector, transaction energy mismatch value embedding coding vector and energy transmission space embedding coding vector; Perform cross-domain query interaction coding on the total revenue embedding coding vector, dual variable embedding coding vector, transaction energy mismatch quota embedding coding vector and energy transmission space embedding coding vector to obtain the multimodal collaborative self-interaction representation vector of economic allocation quota; Generate the user's economic allocation quota value based on the multimodal collaborative self-interaction representation vector of the economic allocation quota; Dispatching is carried out based on the user's economic allocation quota value to achieve optimized grid dispatching.

2. The power grid optimization dispatching method based on intelligent optimization algorithm according to claim 1 is characterized in that: The cross-domain query interactive coding of the total revenue embedding coding vector, the dual variable embedding coding vector, the transaction energy mismatch quota embedding coding vector and the energy transmission space embedding coding vector is performed to obtain the multimodal collaborative self-interaction representation vector of the economic allocation quota, including: The total revenue embedding coding vector is used as the query feature vector, and the feature matrix composed of the total revenue embedding coding vector, the dual variable embedding coding vector, the transaction energy mismatch quota embedding coding vector and the energy transmission space embedding coding vector is used as the key matrix. The query feature vector and the key matrix are input into the cross-domain feature gating query network with local correlation enhanced by the key matrix to obtain the multimodal collaborative self-interaction representation vector of the economic allocation quota.

3. The power grid optimization dispatching method based on intelligent optimization algorithm according to claim 2 is characterized in that: The method uses the total revenue embedding coding vector as the query feature vector, uses the feature matrix composed of the total revenue embedding coding vector, the dual variable embedding coding vector, the transaction energy mismatch quota embedding coding vector and the energy transmission space embedding coding vector as the key matrix, inputs the query feature vector and the key matrix into the cross-domain feature gating query network with local correlation enhancement of the key matrix to obtain the multi-modal collaborative self-interaction representation vector of the economic allocation quota, including: Performing linear transformation on each row vector in the key matrix to obtain an initial key matrix; Calculating feature distribution difference energy coefficients between the query feature vector and each row vector in the initial key matrix to obtain a set of feature distribution difference energy coefficients; Comparing each feature distribution difference energy coefficient in the set of feature distribution difference energy coefficients with a predetermined threshold, and in response to the feature distribution difference energy coefficient being less than the predetermined threshold, taking a mean vector between a row vector corresponding to the feature distribution difference energy coefficient and the query feature vector as an updated row vector to obtain an updated key matrix; Processing the query feature vector using the value embedding matrix to obtain a value vector; Using the query feature vector as the query vector and each row vector in the update key matrix as the key vector, the query vector, the key vector and the value vector are input into a cross-domain attention query module based on a Transformer structure to obtain a sequence of cross-domain query attention vectors of economic allocation quotas; The positional mean vector of the sequence of the cross-domain query focus vector of the economic allocation quota is calculated to obtain a multimodal collaborative self-interaction representation vector of the economic allocation quota.

4. The power grid optimization dispatching method based on intelligent optimization algorithm according to claim 3 is characterized in that: The step of calculating the characteristic distribution difference energy coefficients between the query characteristic vector and each row vector in the initial key matrix to obtain a set of characteristic distribution difference energy coefficients includes: Extracting semantic differences between the query feature vector and each row vector in the initial key matrix respectively to obtain a set of semantically related difference feature vectors; The energy coefficient of each semantic association difference feature vector in the set of semantic association difference feature vectors is calculated to obtain a set of feature distribution difference energy coefficients.

5. The power grid optimization dispatching method based on intelligent optimization algorithm according to claim 4 is characterized in that: The step of respectively extracting the semantic differences between the query feature vector and each row vector in the initial key matrix to obtain a set of semantically associated difference feature vectors includes: Calculate the position point addition, position point subtraction and position point multiplication between the query feature vector and the row vector to obtain a first matching fusion result, a second matching fusion result and a third matching fusion result; After the first matching fusion result, the second matching fusion result and the third matching fusion result are cascaded into a multi-scale semantic association matching vector, one-dimensional convolution processing and maximum pooling processing are performed on the multi-scale semantic association matching vector to obtain a semantic association difference feature vector.

6. The power grid optimization dispatching method based on intelligent optimization algorithm according to claim 5 is characterized in that: The step of calculating the energy coefficient of each semantic association difference feature vector in the set of semantic association difference feature vectors to obtain a set of feature distribution difference energy coefficients includes: The square of the difference between the maximum eigenvalue and the eigenmean of the semantic association difference feature vector, the feature variance of the semantic association difference feature vector, and the bias term are weightedly summed to obtain a weighted sum value, and the sum of the feature variance and the bias term of the semantic association difference feature vector is divided by the weighted sum value to obtain the feature distribution difference energy coefficient.

7. The power grid optimization dispatching method based on intelligent optimization algorithm according to claim 1 is characterized in that: The method of generating the economic allocation quota value of the user based on the multimodal collaborative self-interaction representation vector of the economic allocation quota includes: The multimodal collaborative self-interaction representation vector of the economic allocation quota is input into the decoder to obtain the decoded value of the user's economic allocation quota value.

8. The power grid optimization dispatching method based on intelligent optimization algorithm according to claim 7 is characterized in that: The step of inputting the multimodal collaborative self-interaction representation vector of the economic allocation quota into a decoder to obtain a decoded value of the economic allocation quota value of the user includes: The multimodal collaborative self-interaction representation vector of the economic allocation quota is input into the decoder for decoding calculation, and the multi-layer neural network structure of the decoder model is used to perform layer-by-layer feature extraction and nonlinear mapping on the multimodal collaborative self-interaction representation vector of the economic allocation quota to capture the multi-source data association information contained therein, and generate a decoded value of the economic allocation quota value of the user according to the user's needs and constraints in the economic dispatch of the power grid.

9. A power grid optimization dispatching system based on an intelligent optimization algorithm, based on a power grid optimization dispatching method based on an intelligent optimization algorithm as claimed in any one of claims 1 to 8, characterized in that: include: A smart grid energy model building module is used to build a smart grid energy model; A total benefit value set determination module, used to determine the total benefit value set of each user according to the smart grid energy model; The economic allocation quota value determination module is used to determine the economic allocation quota value of each user based on the total income value set of each user.

10. The power grid optimization dispatching system of the intelligent optimization algorithm according to claim 9 is characterized in that: The economic allocation quota value determination module includes: A data collection unit, used to obtain a user's total revenue set, a dual variable set, a transaction energy mismatch value set, and an energy transmission space variable set; An embedding coding unit, used for embedding the user's total benefit set, dual variable set, transaction energy mismatch amount value set and energy transmission space variable set using an embedding matrix to obtain a total benefit embedding coding vector, a dual variable embedding coding vector, a transaction energy mismatch amount embedding coding vector and an energy transmission space embedding coding vector; A cross-domain query interaction coding unit, used for performing cross-domain query interaction coding on the total revenue embedding coding vector, the dual variable embedding coding vector, the transaction energy mismatch quota embedding coding vector and the energy transmission space embedding coding vector to obtain a multi-modal collaborative self-interaction representation vector of the economic allocation quota; The user's economic allocation quota value determination unit is used to generate the user's economic allocation quota value based on the economic allocation quota multimodal collaborative self-interaction representation vector.

Images