Smart grid adaptive closed-loop power flow optimization control method

Abstract

The application discloses a smart grid adaptive closed-loop power flow optimization control method and relates to the technical field of power grid optimization control.The method comprises the following steps: S1, collecting data to build a dynamic power flow optimization scheme library; S2, extracting characteristic parameters and mapping them into a power grid power flow characteristic mapping matrix; S3, combining real-time working conditions to set dynamic weighting rules, constructing a BP neural network model iteration optimization matrix; S4, matching the optimization matrix with the scheme library similarity according to a preset rule, screening and sorting to determine an optimal power flow optimization scheme; S5, combining historical parameter deviation compensation, debugging power flow control equipment in stages, collecting power flow change values after debugging; and S6, presetting a power flow safety threshold to perform safety verification, dynamically adjusting the scheme according to the result, and forming a closed-loop optimization.The method avoids the problems of one-sided data acquisition and poor adaptability of power grid real-time working conditions by acquiring multi-source core data of the smart grid.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of power grid optimization control, in particular, to a smart grid adaptive closed-loop power flow optimization control method. BACKGROUND

[0002] In the era of deep integration of smart grids and multi-source data fusion technology, power flow fluctuation, as a core regulation and control problem in smart grid operation, has a growing demand for dynamic optimization control. Power grid operators increasingly demand accuracy, operating condition adaptability, and operational economy of power flow control. Multi-source data fusion technology, with its ability to integrate heterogeneous data from the power generation side, transmission side, and user side, has unique advantages in power flow optimization. By integrating power flow fluctuation parameters, real-time operating condition data, and historical control experience, a multi-dimensional dynamic optimization system is formed.

[0003] The core significance of multi-source data fusion-driven power flow optimization is to break through the limitations of traditional single-source data, achieve higher control accuracy, faster response speed, and lower power loss dynamic regulation, and utilize the complementarity of power generation side output fluctuation, transmission side loss characteristics, and user side load changes. Multi-source data fusion technology covers the dynamic needs of each link and solves the problem of regulation lag caused by single data in traditional optimization.

[0004] However, the existing smart grid adaptive closed-loop power flow optimization control method lacks unified multi-source data correlation rules and operating condition adaptation mechanisms, and does not establish a precise mapping relationship between power flow fluctuation characteristics, real-time operating conditions, and control parameters, resulting in poor adaptability of the optimization scheme to real-time operating conditions of the power grid. At the same time, it lacks a standardized process for multiple links, and the safety threshold selection of power flow control parameters is relatively single, making it difficult to fully avoid power flow exceeding limits and equipment overload risks.

[0005] Currently, there is no effective solution to the problems in the related art. SUMMARY

[0006] To overcome the above technical problems existing in the prior art, the present application proposes a smart grid adaptive closed-loop power flow optimization control method.

[0007] To achieve the above purpose, the specific technical solutions adopted by the present application are as follows:

[0008] The smart grid adaptive closed-loop power flow optimization control method comprises the following steps:

[0009] S1, acquire the multi-source core data of the power generation side, power transmission side and user side of the smart grid, real-time working condition parameters and historical tide flow control parameters of the power grid, and establish a smart grid dynamic tide flow optimization scheme library, wherein the historical optimization scheme in the smart grid dynamic tide flow optimization scheme library is one-to-one corresponding associated with the feature matrix;

[0010] S2, extract tide flow fluctuation characteristic parameters in the multi-source core data, preset multi-source data fusion association rules, and map the tide flow fluctuation characteristic parameters to a power grid tide flow characteristic mapping matrix according to the multi-source data fusion association rules;

[0011] S3, set multi-dimensional dynamic weighting rules based on the real-time working condition parameters of the power grid, construct an optimization model using a BP neural network algorithm, and iteratively optimize and adjust the power grid tide flow characteristic mapping matrix to obtain an optimized tide flow characteristic mapping matrix, with the dual optimization objectives of minimizing power loss and reducing tide flow overrun probability.

[0012] S4, preset a dynamic tide flow scheme matching rule, and perform feature similarity matching between the optimized tide flow characteristic mapping matrix and the smart grid dynamic tide flow optimization scheme library based on the dynamic tide flow scheme matching rule, to screen and sort to obtain a smart grid dynamic tide flow optimization scheme;

[0013] S5, combine the smart grid dynamic tide flow optimization scheme with the deviation compensation logic of the historical tide flow control parameters, and perform phased debugging of the smart grid tide flow control equipment, and real-time collection of the power grid tide flow change value after debugging.

[0014] S6, preset a tide flow change safety threshold, perform safety verification on the power grid tide flow change value based on the tide flow change safety threshold, and dynamically adjust the smart grid dynamic tide flow optimization scheme according to the safety verification result to form a closed-loop optimization.

[0015] Preferably, in step S1, the multi-source core data is acquired, including: collecting output fluctuation rate and power stability coefficient from the power generation side through a new energy monitoring system and a unit monitoring platform, collecting line load rate and voltage deviation rate from the power transmission side through a SCADA system and a WAMS system, and collecting load mutation rate and power consumption time period distribution coefficient from the user side through a smart meter and a load monitoring terminal.

[0016] Preferably, in step S2, the tide flow fluctuation characteristic parameters include output fluctuation rate, power stability coefficient, line load rate, voltage deviation rate, load mutation rate and power consumption time period distribution coefficient.

[0017] Preferably, in the step S2, the preset multi-source data fusion correlation rule comprises: setting the power generation side data weight as 40%, the power transmission side data weight as 30%, and the user side data weight as 30%, and configuring initial weights for the characteristic parameters of the power flow fluctuation of each side, wherein the power generation side output fluctuation rate weight is 0.25, the power stability coefficient weight is 0.15, the power transmission side line load rate weight is 0.2, the voltage deviation rate weight is 0.1, the user side load mutation rate weight is 0.15, and the power consumption time interval distribution coefficient weight is 0.15, and the sum of the initial weights of all the characteristic parameters is 1.

[0018] Preferably, in the step S3, the BP neural network algorithm is used to construct the optimization model, which comprises: constructing a three-layer structure of an input layer, a hidden layer and an output layer, wherein the number of input layer nodes is the total number of elements of the power grid power flow characteristic mapping matrix, the number of output layer nodes is 2 and corresponds to the power loss prediction value and the power flow overrun probability prediction value respectively, and the number of hidden layer nodes is determined based on the number of input layer nodes and the number of output layer nodes through an empirical formula.

[0019] Preferably, the empirical formula is:

[0020] ;

[0021] wherein n is the number of hidden layer nodes, m is the number of input layer nodes, k is the number of output layer nodes, and a is an adjustment constant between 10 and 20.

[0022] Preferably, in the step S4, the dynamic power flow scheme matching rule is a cosine similarity algorithm, and the formula of the cosine similarity algorithm is:

[0023] ,

[0024] wherein A is the optimized power flow characteristic mapping matrix, and B is the characteristic matrix of the historical optimization scheme in the scheme library.

[0025] Preferably, the preset characteristic similarity screening threshold is 0.85, and when the calculated characteristic similarity is greater than or equal to 0.85, the corresponding historical optimization scheme is included in the candidate optimization scheme set.

[0026] Preferably, in the step S5, the smart grid power flow control device is debugged in stages, which comprises: a pre-debugging stage, a fine-tuning debugging stage and a stable verification stage.

[0027] Preferably, in the step S6, the preset power flow change safety threshold is set based on the power grid safe operation standard and the historical stable working condition data, and the power flow change safety threshold comprises a power loss safety interval, a power flow overrun probability safety interval and a voltage frequency fluctuation safety interval.

[0028] The present application has the following advantages:

[0029] 1、The application acquires the multi-source core data of the smart grid, the real-time working condition parameters and the historical tide flow control parameters of the grid, presets the dynamic tide flow optimization scheme library of the smart grid, avoids the problems of one-sided data acquisition and poor adaptability of the control scheme to the real-time working condition in the traditional grid tide flow optimization, reduces the risk of excessive power loss or tide flow exceeding limit caused by fixed parameter matching, and through the corresponding correlation between the tide flow fluctuation characteristic parameters and the tide flow control core parameters of the power generation side, the power transmission side and the user side, builds a multi-source data dynamic weight fusion rule, ensures that the adaptation logic of the tide flow characteristics and the control parameters is consistent with the operation principle of the multi-source heterogeneous data of the smart grid, avoids invalid parameter combination interference, and improves the accuracy of the tide flow characteristic mapping matrix construction.

[0030] 2、The application builds a three-level optimization system of weight dynamic adjustment, matrix weighting correction and abnormal data elimination, eliminates the risk parameters that exceed the safe operation range of the grid, solves the pain points of ignoring the data fluctuation difference and individual working condition characteristics of each side, improves the safety and individual adaptation degree of the tide flow optimization scheme, reduces the equipment debugging cost and time through historical parameter reuse, reduces the power grid operation loss and improves the tide flow operation stability, so that the scheme has more value and scale promotion significance in the power system.

[0031] 3、The application matches the preset dynamic tide flow scheme matching rule, combines multi-dimensional weighted operation and feature similarity screening, realizes hierarchical matching of the optimization scheme, selects a safe candidate scheme set from the optimized feature matrix based on the safety priority rule, and determines the safety bottom line of the tide flow control, and then through the optimal effect rule and the historical adaptation rule, the safe candidate scheme set and the scheme library are accurately matched, the reliability of the matching result is ensured through the iterative optimization verification mechanism, and the optimization deviation caused by a single rule is avoided; at the same time, the optimized scheme after matching is combined with the historical tide flow control parameter to debug the grid tide flow control equipment, realizes the whole process closed loop, and improves the operability of the optimization scheme. BRIEF DESCRIPTION OF DRAWINGS

[0032] In order to more clearly illustrate the technical schemes in the embodiments of the present application or the prior art, the drawings needed in the embodiments will be briefly introduced as follows. Obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can be obtained by those skilled in the art without creative labor.

[0033] Figure 1 It is a method flow chart of the smart grid adaptive closed-loop tide flow optimization control method according to the embodiment of the present application. DETAILED DESCRIPTION

[0034] The specific embodiments of the present application will be further described in detail below with reference to the accompanying drawings and examples. The following examples are used to illustrate the present application, but are not intended to limit the scope of the present application.

[0035] Therefore, the detailed description of the embodiments of the present application provided below in the accompanying drawings is not intended to limit the scope of the claimed application, but only represents selected embodiments of the present application. All other embodiments obtained by those of ordinary skill in the art based on the embodiments in the present application without creative labor are within the scope of protection of the present application.

[0036] The present application will be further described in conjunction with the accompanying drawings and specific embodiments, as shown in the drawings, the smart grid adaptive closed-loop power flow optimization control method according to the embodiments of the present application comprises the following steps: Figure 1

[0037] S1, obtaining the multi-source core data of the generation side, transmission side and user side of the smart grid, the real-time working condition parameters of the power grid and the historical power flow control parameters, and establishing a smart grid dynamic power flow optimization scheme library, and the historical optimization schemes in the smart grid dynamic power flow optimization scheme library are one-to-one corresponding associated with the feature matrix;

[0038] Specifically, the generation side collects data such as output fluctuation rate and power stability coefficient through a new energy monitoring system and a unit monitoring platform; the transmission side obtains parameters such as line load rate and voltage deviation rate relying on a SCADA system and a WAMS system; and the user side collects information such as load mutation rate and power consumption time period distribution coefficient through a smart meter and a load monitoring terminal.

[0039] The real-time working condition parameters (voltage, frequency, load rate) of the power grid are collected in real time through a global sensing network, and the historical power flow control parameters are called from a power grid dispatching historical database, including records such as equipment debugging deviation and control response delay. The scheme library needs to integrate previous effective optimization schemes, extract the corresponding power flow feature mapping matrix (generated by multi-source data fusion) of each scheme, and establish a one-to-one corresponding relationship between the historical optimization scheme and the feature matrix through an indexing mechanism, to ensure that the scheme library can quickly match and adapt the scheme according to the current working condition feature matrix.

[0040] S2, extracting the power flow fluctuation feature parameters in the multi-source core data, presetting a multi-source data fusion association rule, and the core of the multi-source data fusion association rule is to match the actual influence of each side data weight distribution on the power flow fluctuation, and mapping the power flow fluctuation feature parameters to a power grid power flow feature mapping matrix according to the association rule;

[0041] ​In the embodiment of the present application, the power flow fluctuation characteristic parameters in the multi-source core data are extracted, a multi-source data fusion association rule is preset, and the core of the multi-source data fusion association rule is the matching of the weight distribution of each side data to the actual influence on the power flow fluctuation. According to the association rule, the power flow fluctuation characteristic parameters are mapped into a power grid power flow characteristic mapping matrix, which includes the following steps:

[0042] S21, classifying and preprocessing the multi-source core data, eliminating abnormal values and completing missing values to obtain power generation side data, power transmission side data and user side data;

[0043] Specifically, first, divide into three categories according to data sources: power generation side, power transmission side and user side. The power generation side data includes new energy output, conventional unit operation parameters, etc. The power transmission side covers line active / reactive power flow, voltage deviation, etc. The user side includes industrial and residential load, electricity time period distribution, etc. Then, perform format standardization and time alignment processing to unify data dimension and sampling frequency.

[0044] The 3σ principle is used to eliminate abnormal values, and data exceeding ±3 times the standard deviation of the mean is removed. For missing values, short-term missing values are completed by linear interpolation, and long-term missing values are predicted and filled based on similar working condition historical data through KNN algorithm. Through data consistency verification, it is ensured that there is no logical conflict in the three side data, forming clean and regular power generation side, power transmission side and user side data sets, laying a foundation for subsequent power flow fluctuation characteristic parameter extraction.

[0045] S22, extracting the power flow fluctuation characteristic parameters corresponding to the power generation side data, the power transmission side data and the user side data, and the power flow fluctuation characteristic parameters include output fluctuation rate, line load rate, load mutation rate, power stability coefficient, voltage deviation rate and electricity time period distribution coefficient;

[0046] Specifically, based on the clean data of the power generation side, the power transmission side and the user side after classification preprocessing, 6 types of power flow fluctuation characteristic parameters are extracted according to the side orientation: the power generation side calculates the output fluctuation rate by calculating the output change amplitude in unit time through new energy output and conventional unit operation data, and calculates the power stability coefficient based on the standard deviation and mean proportion of the output data; the power transmission side determines the line load rate based on the ratio of real-time active / reactive power flow data to thermal stability limit capacity of the line, and extracts the voltage deviation rate by the difference between the actual voltage and the rated voltage; the user side calculates the load mutation rate by calculating the ratio of load mutation amplitude to reference load in a short time based on real-time monitoring data of load, and generates the electricity time period distribution coefficient by statistical electricity load proportion in different time periods. In the extraction process, the rationality of the parameters is verified synchronously, and the abnormal values exceeding the normal operation range of the power grid are removed to ensure that the characteristic parameters accurately reflect the power flow fluctuation characteristics of each side, laying a foundation for subsequent multi-source data fusion and mapping matrix construction.

[0047] S23, set the power generation side data weight 40%, the power transmission side data weight 30%, the user side data weight 30%, form a multi-source data fusion association rule, and calculate the power flow fluctuation comprehensive value of each dimension feature based on the multi-source data fusion association rule using a weighted summation algorithm;

[0048] In the embodiment of the application, the setting of the power generation side data weight 40%, the power transmission side data weight 30%, and the user side data weight 30% forms a multi-source data fusion association rule, and the calculation of the power flow fluctuation comprehensive value of each dimension feature based on the multi-source data fusion association rule using a weighted summation algorithm includes the following steps:

[0049] S231, it is clear that the core of the multi-source data fusion association rule is that the data weight of the power generation side, the power transmission side and the user side is positively correlated with the influence degree of each side on the power flow fluctuation of the power grid, and the reference weight proportion is 40% for the power generation side, 30% for the power transmission side and 30% for the user side;

[0050] Specifically, taking “the positive correlation between the data weight of each side and the influence degree on the power flow fluctuation of the power grid” as the core principle, combined with the direct dominant role of the power generation side output fluctuation on the power flow, the intermediate influence of the transmission characteristics of the power transmission side, and the relatively moderate influence of the user side load change, the reference weight proportion is set as 40% for the power generation side, 30% for the power transmission side and 30% for the user side. The initial weight of each side power flow fluctuation characteristic parameter is configured (such as power generation side output fluctuation rate 0.25, power transmission side line load rate 0.2, etc.), and the fluctuation threshold is set according to the characteristics of each parameter. The weight is adjusted dynamically (power is increased when the threshold is exceeded, and the weight is reduced by 50% when the threshold is lower than the threshold), and the weight sum is ensured to be 1±0.01 through summation verification. This rule not only anchors the objective basis of weight allocation, but also adapts to the change of power grid working condition through dynamic adjustment, laying a foundation for multi-source data accurate fusion and power flow characteristic mapping matrix construction.

[0051] S232, the initial weight of each side power flow fluctuation characteristic parameter is configured, wherein the power generation side output fluctuation rate is 0.25, the power stability coefficient is 0.15, the power transmission side line load rate is 0.2, the voltage deviation rate is 0.1, the user side load mutation rate is 0.15, and the power consumption time period distribution coefficient is 0.15, and the sum of the initial weights of all characteristic parameters is 1;

[0052] In the embodiment of the application, the initial weight of each side power flow fluctuation characteristic parameter is configured, wherein the power generation side output fluctuation rate is 0.25, the power stability coefficient is 0.15, the power transmission side line load rate is 0.2, the voltage deviation rate is 0.1, the user side load mutation rate is 0.15, and the power consumption time period distribution coefficient is 0.15, and the sum of the initial weights of all characteristic parameters is 1 includes the following steps:

[0053] S2321. Based on the characteristics of the power flow fluctuation parameters on each side, fluctuation thresholds are set respectively, including power generation side output fluctuation rate ±10%, power stability coefficient ±5%, transmission side line load rate ±8%, voltage deviation rate ±3%, user side load mutation rate ±12%, and electricity consumption period distribution coefficient ±6%.

[0054] Specifically, based on the impact of each parameter on power grid flow stability, actual operating characteristics, and equipment capacity, differentiated threshold standards are established. For generation-side power output, significantly affected by new energy fluctuations, a power output fluctuation rate of ±10% is set to adapt to dynamic changes in wind and solar power output, and a power stability coefficient of ±5% is set to strictly control the stability of generation-side power output. For the transmission side, directly related to line safety operation, a line load rate threshold of ±8% is set to align with the line's thermal stability limit, and a voltage deviation rate of ±3% is limited to meet grid voltage quality requirements. For the user-side, load variations are diverse and highly random; a load mutation rate of ±12% is set to accommodate sudden fluctuations in industrial and residential loads, and a power consumption time distribution coefficient of ±6% is set to accommodate reasonable deviations in power load at different times. This threshold system not only conforms to the operating patterns of parameters on each side but also provides clear judgment benchmarks for subsequent dynamic weight adjustments and power flow optimization control, ensuring the safety and flexibility of grid operation.

[0055] S2322. Compare the actual fluctuation amplitude of each feature parameter with the corresponding fluctuation threshold side by side. If the actual fluctuation amplitude exceeds the threshold, increase the weight of the feature parameter by a step size of 0.05. If the actual fluctuation amplitude is less than 50% of the threshold, decrease the weight of the feature parameter by a step size of 0.03.

[0056] Specifically, based on preset thresholds for power flow fluctuation characteristic parameters on each side (e.g., power generation fluctuation rate ±10%, transmission line load rate ±8%, user load mutation rate ±12%), a precise comparison is made between the actual fluctuation amplitude and the corresponding threshold on each side. If the actual fluctuation amplitude of a certain characteristic parameter exceeds the preset threshold, the weight of that parameter is increased by a step of 0.05 to strengthen its impact on power flow optimization; if the actual fluctuation amplitude is less than 50% of the threshold, the corresponding weight is decreased by a step of 0.03 to reduce interference from non-critical parameters.

[0057] After the weights are adjusted, a total verification is required to ensure that the total weights of all feature parameters are 1 ± 0.01. If this is not met, the weights of the user-side parameters are fine-tuned first (adjustment step size 0.01). The dynamic adjustment logic not only fits the operating characteristics of each side parameter, but also adapts to the real-time changes in the power grid operating conditions, thereby improving the pertinence and accuracy of multi-source data fusion.

[0058] S2323. Perform a total check on all adjusted feature parameter weights to ensure that the total weight is 1 ± 0.01. If the condition is not met, prioritize adjusting the user-side feature parameter weights.

[0059] Specifically, the adjusted weights of all feature parameters on the power generation side, power transmission side and user side are first summarized, the total value is calculated and compared with the standard range of 1±0.01. If the total value exceeds the range, the weights of the user side feature parameters (load mutation rate, power consumption time period distribution coefficient) are preferentially fine-tuned, and the adjustment step is set to 0.01 to avoid significantly changing the rationality of the weight distribution of other sides. After fine-tuning, the weight total is calculated again, and the iteration is repeated until the requirement of 1±0.01 is met. This checking mechanism can ensure the self-consistency of the weight distribution logic of the multi-source data fusion association rule, guarantee the accuracy of the subsequent weighted summation algorithm, and provide a reliable weight basis for the construction of the power flow feature mapping matrix.

[0060] S2324, store the adjusted weight to the weight parameter library and use it as the initial weight reference under the same working condition in the future.

[0061] Specifically, the adjusted weight that passes the total check (satisfies 1±0.01) is bound with the corresponding working condition feature label (including power generation side / power transmission side / user side fluctuation parameters, real-time working conditions such as grid voltage / frequency / load rate, peak-valley period information) to form an “weight data-working condition feature” association data group. According to the working condition type (such as high load / high penetration of new energy / stable operation), time period and other dimensions, it is stored in the weight parameter library and an index directory is established for quick retrieval. At the same time, a data updating mechanism is set to cover old data or supplement records with the latest weight data adapted to the working condition. When encountering the same / similar grid working condition in the future, the corresponding weight is retrieved from the parameter library through working condition feature matching and directly used as the initial weight reference, without the need to repeat the “threshold comparison-weight adjustment” process, which reduces the calculation time and guarantees the weight adaptability, improving the efficiency of multi-source data fusion and power flow optimization.

[0062] S233, calculate the power flow fluctuation comprehensive value by using the weighted summation algorithm, wherein the formula of the weighted summation algorithm is:

[0063] F=0.4×(F1×W1+F2×W2)+0.3×(F3×W3+F4×W4)+0.3×(F5×W5+F6×W6);

[0064] Wherein, F is the power flow fluctuation comprehensive value, F1-F6 are the specific values of each feature parameter, and W1-W6 are the initial weights of the corresponding feature parameters;

[0065] S234, perform effectiveness check on the power flow fluctuation comprehensive value, eliminate abnormal comprehensive values that exceed the reasonable range of the normal operation scenario of the power grid, and retain the effective comprehensive values for constructing the initial feature matrix.

[0066] Specifically, based on the historical data of the normal operation scene of the power grid and the industry standard, the reasonable range of the comprehensive value is determined as 0-0.8, which is suitable for the normal fluctuation characteristics of the data on the power generation side, the power transmission side and the user side. The range of the comprehensive value obtained by the weighted summation algorithm is checked one by one, and the abnormal values exceeding the range of 0-0.8 are directly removed to avoid interference of extreme data in the matrix construction. At the same time, combined with the multi-source data fusion correlation rule, the consistency of the comprehensive value and the corresponding power grid working condition is checked, and if the comprehensive value and the real-time working condition parameter are logically conflicted (such as high fluctuation comprehensive value in a low load period), it is also determined as abnormal and removed. Finally, all the effective comprehensive values meeting the range and the logical consistency are collected, and according to the correspondence between the power flow fluctuation characteristic parameter type and the power grid operation time node, clean and reliable data support is provided for the initial feature matrix construction.

[0067] S24, constructing an initial feature matrix with the power flow fluctuation characteristic parameter type as the row and the power grid operation time node as the column, combining the multi-source data fusion correlation rule to weight and correct each element in the initial feature matrix, and after dimension checking and abnormal element removal, obtaining a power grid power flow characteristic mapping matrix.

[0068] In the embodiments of the present application, the initial feature matrix is constructed with the power flow fluctuation characteristic parameter type as the row and the power grid operation time node as the column, the multi-source data fusion correlation rule is combined to weight and correct each element in the initial feature matrix, and after dimension checking and abnormal element removal, the power grid power flow characteristic mapping matrix is obtained, including the following steps:

[0069] S241, taking the 6 types of power flow fluctuation characteristic parameters as the row vector and the 24-hour continuous operation time node of the power grid as the column vector, filling the effective power flow fluctuation comprehensive value according to the correspondence between the characteristic parameter and the time node to construct an initial feature matrix;

[0070] Specifically, taking the 6 types of power flow fluctuation characteristic parameters of output fluctuation rate, line load rate, load mutation rate, power stability coefficient, voltage deviation rate and power consumption time distribution coefficient as the matrix row vector, and taking each time node of the 24-hour continuous operation of the power grid as the column vector, the one-to-one correspondence of the "characteristic parameter-time node" is determined. The power flow fluctuation comprehensive value that has passed the effectiveness check (removed from the reasonable range of 0-0.8) is accurately filled into the corresponding position of the matrix according to the type of the characteristic parameter and the collection time node, and an initial feature matrix with a dimension of 6 rows x 24 columns is constructed. The data consistency is checked synchronously during the filling process to ensure that the comprehensive values of each characteristic parameter under the same time node come from the same power grid operation scene, and the 24 time node data of the same characteristic parameter are continuous and complete, laying a regular data foundation for subsequent matrix weight correction and power grid power flow characteristic mapping matrix construction.

[0071] S242, determine the correction weight according to the multi-source data fusion correlation rule, wherein the generation side characteristic parameter corresponds to the matrix element correction weight 1.2, the transmission side corresponds to the element 1.0, and the user side corresponds to the element 0.8, and the correction weight is consistent with the influence degree of each side data on the power flow;

[0072] Specifically, according to the rule that the correction weight is strongly related to the influence degree of each side data on the power flow fluctuation, and combining the actual action difference of each side data, the weight is allocated. The generation side output fluctuation directly dominates the power flow change, has the strongest influence on the power flow stability, and the corresponding matrix element correction weight is set to 1.2; the transmission side bears the function of power transmission, and the influence degree is moderate, and the correction weight is set to 1.0; the user side load fluctuation is relatively moderate, and the indirect influence on the power flow is weak, and the correction weight is set to 0.8. When the weight is allocated, according to the generation side, the transmission side and the user side category of each element in the initial characteristic matrix, the corresponding correction weight is matched one by one, so that the matrix element correction degree of each type of characteristic parameter is matched with the actual influence, which not only conforms to the core of the multi-source data fusion correlation rule, but also strengthens the accuracy of the power flow characteristic mapping through the differentiated weight, and provides a reliable basis for subsequent matrix optimization.

[0073] S243, the calculation method of multiplying the element original value by the corresponding correction weight is adopted, and all elements in the initial characteristic matrix are weighted and corrected point by point to obtain an intermediate characteristic matrix;

[0074] Specifically, first, according to the multi-source data fusion correlation rule, the correction weight is matched according to the side to which the element belongs (the generation side characteristic parameter corresponds to the element 1.2, the transmission side corresponds to the element 1.0, and the user side corresponds to the element 0.8), so that the correction degree is consistent with the influence degree of each side data on the power flow. Then, all elements in the initial characteristic matrix (6 rows x 24 columns) are traversed, and the calculation of “original value x corresponding correction weight” is performed on each element one by one to accurately strengthen the power flow influence weight of the key side data. After the calculation is completed, all the corrected elements are integrated according to the “characteristic parameter-time node” position of the original matrix, to form a structured intermediate characteristic matrix, which provides accurate corrected basic data for subsequent dimension checking and abnormal element elimination, and improves the construction reliability of the power flow characteristic mapping matrix.

[0075] S244, dimension checking and abnormal element elimination are performed on the intermediate characteristic matrix to obtain a power flow characteristic mapping matrix.

[0076] Specifically, the dimension of the intermediate feature matrix is first checked, and it is verified whether the dimension of the core check matrix is 6 rows x 24 columns to ensure complete correspondence with 6 types of power flow fluctuation characteristic parameters and 24-hour time nodes, and the matrix dimension is also verified for adaptability to the current power grid operation node quantity. If the dimensions do not match, the missing columns or redundant columns are supplemented or pruned to ensure compliance of the matrix structure. Then, abnormal elements are removed, and a reasonable threshold of the matrix elements under normal operation of the power grid is set to be less than or equal to 1.0. The element values are screened one by one, and abnormal values exceeding the threshold are removed. For the gaps caused by the removal, the neighboring interpolation method is used to complete the missing data with the mean value of the effective elements of the same characteristic parameter of the adjacent time nodes, so as to ensure data continuity. After completing the dimension check and abnormality removal, the matrix data is verified for consistency to ensure no logical conflicts, and finally the power grid power flow feature mapping matrix that is accurate, complete and adaptive to the power grid operating conditions is formed, thereby providing reliable data support for subsequent power flow optimization modeling.

[0077] S3, based on the real-time operating parameter of the power grid, a multi-dimensional dynamic weighting rule is set, a BP neural network algorithm is used to construct an optimization model, and the power grid power flow feature mapping matrix is iteratively optimized and adjusted with the dual optimization targets of minimizing the power loss of the power grid and reducing the probability of power flow exceeding the limit;

[0078] In the embodiments of the present application, the multi-dimensional dynamic weighting rule based on the real-time operating parameter of the power grid is set, the BP neural network algorithm is used to construct an optimization model, and the power grid power flow feature mapping matrix is iteratively optimized and adjusted with the dual optimization targets of minimizing the power loss of the power grid and reducing the probability of power flow exceeding the limit, which includes the following steps:

[0079] S31, the real-time operating parameters of the power grid are normalized, abnormal parameters exceeding the reasonable operating range are removed, and a standardized operating parameter set is obtained. The voltage, frequency and load rate are taken as the core dimensions, and a multi-dimensional dynamic weighting rule is set with the weight dynamically floating according to the peak and valley period;

[0080] Specifically, the real-time operating parameters of the power grid are first collected, including voltage, frequency, load rate and auxiliary operating parameters. The min-max standardization is used to map all parameters to the [0, 1] interval to eliminate the dimension difference and ensure the comparability of the parameters. Then, the reasonable operating range of each parameter is set, and abnormal data such as voltage exceeding the limit, frequency exceeding the standard and load rate overload are removed. After verification, a clean standardized operating parameter set is obtained. The voltage, frequency and load rate are selected as the core dimensions for regulation and control. A multi-dimensional dynamic weighting rule is set based on the characteristics of the peak and valley periods of the power grid. The load rate weight is increased during the peak load period, the weights of the three are balanced during the flat period, and the voltage stability is emphasized and the weight is increased during the valley period. The dynamic allocation of the weight always matches the real-time operating demand of the power grid, thereby providing standardized and highly adaptive operating data support for subsequent BP neural network optimization modeling, and ensuring that the optimization direction accurately matches the actual operating state of the power grid.

[0081] S32, a BP neural network optimization model is constructed by using a BP neural network algorithm, and the BP neural network optimization model is a three-layer structure of an input layer, a hidden layer and an output layer, input layer node data is an element of a power flow characteristic mapping matrix, output layer node data is a power loss prediction value and a power flow overrun probability prediction value, and the number of hidden layer nodes is determined based on the number of input layer nodes and the number of output layer nodes through an empirical formula;

[0082] In the embodiment of the application, the BP neural network optimization model is constructed by using the BP neural network algorithm, and the BP neural network optimization model is a three-layer structure of an input layer, a hidden layer and an output layer, input layer node data is an element of a power flow characteristic mapping matrix, output layer node data is a power loss prediction value and a power flow overrun probability prediction value, and the number of hidden layer nodes is determined based on the number of input layer nodes and the number of output layer nodes through an empirical formula, including the following steps:

[0083] S321, the number of input layer nodes is the total number of elements of the power flow characteristic mapping matrix, and the number of output layer nodes is 2, and respectively corresponds to the power loss prediction value and the power flow overrun probability prediction value;

[0084] Specifically, the network level node quantity setting logic is first determined, the number of input layer nodes corresponds to the total number of elements of the power flow characteristic mapping matrix of the power grid, the matrix is a 6-row x 24-column structure, so the number of input layer nodes is determined to be 144, ensuring that all power flow characteristic data is completely input, providing comprehensive basic support for model training. According to the double optimization target, two nodes are set in the output layer, one node is specially corresponding to the power loss prediction value of the power grid, the prediction result of the power loss of the power grid is accurately output by the model, and the other node corresponds to the power flow overrun probability prediction value, which directly feeds back the probability of the power flow overrun risk of the power grid, the double output nodes not only meet the core optimization demand of minimizing power loss and reducing power flow overrun probability, but also make the model output target clear and explicit, providing accurate direction for subsequent iteration optimization and result determination, and ensuring that the construction of the BP neural network optimization model meets the actual demand of power flow regulation of the power grid.

[0085] S322, the number of hidden layer nodes is determined by using an empirical formula, wherein the empirical formula is:

[0086] ;

[0087] Wherein, n is the number of hidden layer nodes, m is the number of input layer nodes 144, k is the number of output layer nodes 2, a is an adjustment constant between 10-20, and the number of hidden layer nodes is calculated to be 43;

[0088] S323, the core auxiliary parameters of the BP neural network optimization model are configured, wherein the initial value of the learning rate is 0.03, the maximum number of iterations is 500, and the convergence condition is that the loss function value is less than 0.001;

[0089] Specifically, in combination with the double objectives of power grid power flow optimization (minimizing power loss and reducing power flow over-limit probability) and the training characteristics of 144-dimensional input data, key parameters are accurately set. The initial learning rate is configured as 0.03, which takes into account the convergence speed and training stability of the model, avoiding parameter oscillation caused by excessively high learning rate and excessively long iteration period caused by excessively low learning rate. The maximum number of iterations is set to 500, which is suitable for the dimension size of the power flow feature mapping matrix, balancing training efficiency and optimization depth. The convergence condition is explicitly defined as a loss function value < 0.001, which is used as the core judgment standard for terminating model training. After parameter configuration, small sample pre-training verification is carried out. If convergence is slow or oscillation occurs during pre-training, the learning rate is fine-tuned by 0.01 to ensure that the convergence condition can be met within 500 iterations. The final parameter combination not only fits the training rules of power grid data, but also ensures that the model quickly converges to the optimal state, providing stable and reliable model parameter support for the iterative optimization of the power flow feature matrix.

[0090] S324, set the initial range of connection weight values between the input layer and the hidden layer, and between the hidden layer and the output layer as [-0.5, 0.5], and the activation function uses the Sigmoid function.

[0091] Specifically, when configuring the connection weight values and activation functions of the BP neural network, first set the initial range of the connection weight values between the layers, and the connection weight values between the input layer and the hidden layer and between the hidden layer and the output layer are uniformly set as the interval [-0.5, 0.5]. This interval has a moderate value, which can avoid the saturation of neurons caused by excessively large initial weight values and the disappearance of training gradients caused by excessively small initial weight values, ensuring the stability of the initial training of the model and laying a foundation for subsequent iterative optimization. The activation function uses the Sigmoid function, which can realize the nonlinear mapping of input signals and adapt to the complex nonlinear relationship of power grid power flow optimization. At the same time, it can output results in the interval 0-1, which is highly consistent with the prediction scenarios of power loss and power flow over-limit probability. After configuration, it is verified simultaneously to ensure that the weight values are randomly and uniformly distributed in the set interval, and the gradient of the activation function is in the effective range, avoiding gradient dispersion affecting convergence, so that the model can efficiently process power flow feature mapping matrix data, accurately output double objective prediction values, and improve the nonlinear fitting ability and prediction accuracy of the optimization model.

[0092] S33, preset an iterative optimization judgment rule, and iteratively optimize the power flow feature mapping matrix through the iterative optimization judgment rule. When the iterative optimization judgment rule is met, the optimization is stopped and the matrix is output. When the iterative optimization judgment rule is not met, the proportion of each dimension weight in the dynamic weighting rule is adjusted and then iterated again.

[0093] Specifically, a preset iterative optimization judgment rule is set, and the core is around the double optimization target and the convergence requirement. The rule is clear that the power loss prediction value of the model output is reduced to a preset reasonable threshold, the power flow out-of-limit probability prediction value is lower than a limited value, and the loss function value of continuous 10 iterations is less than 0.001, and the change rate of the power flow characteristic mapping matrix elements between two iterations is less than 0.02. When executing, the matrix is input into the BP neural network model for iterative optimization. After each iteration, it is verified whether the judgment rule is met. If it is met, the optimization is stopped and the optimal matrix is output. If it is not met, the multi-dimensional dynamic weighting rule is adjusted, the core dimension weight proportion of the voltage, frequency and load rate is preferentially fine-tuned, and the peak-valley period adaptive adjustment amplitude is adjusted. After adjustment, the model is input for iteration again. The parameter rationality is monitored synchronously in the iteration to avoid weight imbalance, and the cycle is repeated until the standard is reached, so as to guarantee that the optimization matrix meets the double target demand, and to improve the power flow control accuracy and the power grid operation stability.

[0094] S4, a dynamic power flow scheme matching rule is preset, and the optimized and adjusted power flow characteristic mapping matrix is matched with the intelligent power grid dynamic power flow optimization scheme library based on the dynamic power flow scheme matching rule to screen and sort the intelligent power grid dynamic power flow optimization scheme;

[0095] In the embodiment of the application, the preset dynamic power flow scheme matching rule includes the following steps:

[0096] S41, the preset dynamic power flow scheme matching rule is a cosine similarity algorithm, and the preset candidate scheme priority sorting rule is similarity priority, power loss reduction amplitude priority and debugging cost priority;

[0097] Specifically, the core feature vectors of the power grid power flow characteristic mapping matrix and the historical scheme feature vectors in the dynamic power flow optimization scheme library are extracted first, and the cosine value of the included angle of the two is calculated through the cosine similarity. The closer the value is to 1, the higher the matching degree is. In this way, the candidate scheme suitable for the current working condition is accurately screened. At the same time, the preset candidate scheme priority sorting rule is set, and the schemes are sorted in order of similarity priority, power loss reduction amplitude priority and debugging cost priority. The standard scheme is selected by comparing the power loss reduction amplitude, and the optimal scheme is finally determined by accounting for the historical debugging cost, taking into account the adaptability, economy and practicability. This rule not only ensures that the scheme is accurately adapted to the real-time working condition, but also maximizes the reduction of loss and cost, provides reliable support for subsequent safety verification and closed-loop debugging, and improves the power flow optimization efficiency and practicability.

[0098] S42, standardize the power flow characteristic mapping matrix after the optimization adjustment, so that the power flow characteristic mapping matrix and the characteristic matrix of the historical optimization scheme in the intelligent power grid dynamic power flow optimization scheme library are consistent in dimension;

[0099] Specifically, the power flow characteristic mapping matrix after the optimization adjustment is first standardized. The core target is to make it completely consistent with the dimension of the historical optimization scheme characteristic matrix in the scheme library, to ensure the effectiveness of matching. First, verify the current matrix dimension, and determine the unified standard dimension of the historical characteristic matrix in the scheme library. Then, regularize the dimension of the matrix after optimization. If the dimension is insufficient, supplement the interpolation data of the corresponding time period and characteristic parameters. If the dimension is redundant, eliminate the invalid rows and columns to ensure accurate matching of the number of rows and columns. Then, using min-max standardization, map all elements of the matrix to the [0, 1] interval, eliminate the influence of dimension difference and numerical fluctuation, and align the data accuracy and format standard with the historical matrix. After processing, the consistency is verified synchronously to confirm that there is no dimension deviation and numerical specification, so that the standardized power flow characteristic mapping matrix and the historical scheme characteristic matrix have comparable conditions, laying a unified data foundation for subsequent cosine similarity matching, and improving the accuracy of scheme matching.

[0100] S43, calculate the characteristic similarity between the power flow characteristic mapping matrix after the optimization adjustment and the power grid dynamic power flow optimization scheme in the intelligent power grid dynamic power flow optimization scheme library based on the cosine similarity algorithm, and preset a characteristic similarity filtering threshold. Filter the characteristic similarity based on the characteristic similarity filtering threshold to obtain a candidate optimization scheme set.

[0101] In the embodiment of the application, the characteristic similarity between the power flow characteristic mapping matrix after the optimization adjustment and the power grid dynamic power flow optimization scheme in the intelligent power grid dynamic power flow optimization scheme library is calculated based on the cosine similarity algorithm, and a characteristic similarity filtering threshold is preset. The characteristic similarity is filtered based on the characteristic similarity filtering threshold to obtain a candidate optimization scheme set, which includes the following steps:

[0102] S431, the formula of the cosine similarity algorithm is:

[0103] ;

[0104] Wherein, is the similarity degree of the power flow characteristic mapping matrix after optimization and the characteristic matrix of the historical optimization scheme in the scheme library, A is the power flow characteristic mapping matrix after optimization adjustment, and B is the characteristic matrix of the historical optimization scheme in the scheme library;

[0105] S432, the characteristic similarity filtering threshold is 0.85. When the calculated characteristic similarity is greater than or equal to 0.85, the corresponding historical optimization scheme is included in the candidate optimization scheme set.

[0106] Specifically, the feature similarity screening threshold is preset as 0.85, which is determined in combination with the matching demand of historical adaptation cases and power grid operating conditions, and takes into account the scheme adaptability and screening efficiency. After standardization processing, the power flow feature mapping matrix and the historical optimization scheme feature matrix in the intelligent power grid dynamic power flow optimization scheme library are aligned and calculated for feature similarity value one by one through the cosine similarity algorithm. The calculation results are compared with the threshold value of 0.85. When the feature similarity is greater than or equal to 0.85, it is determined that the historical scheme is highly adaptable to the current power grid operating conditions, and is directly included in the candidate optimization scheme set; if it is less than 0.85, it means that the adaptation degree is insufficient, and it is excluded to avoid invalid schemes increasing the screening cost. The similarity value and the core adaptation point of the qualified scheme are recorded during the screening, and the corresponding historical operating conditions and optimization effectiveness are marked, which provides a clear basis for subsequent priority sorting and ensures that the candidate scheme set has adaptability and practicality, laying a solid foundation for accurately selecting the optimal power flow optimization scheme.

[0107] S433, if the candidate optimization scheme set is empty, the screening threshold is reduced by 0.05 step by step until at least 3 candidate optimization schemes are screened out;

[0108] Specifically, the candidate scheme set supplement mechanism needs to be preset. After completing the feature consistency check, it is checked whether the candidate optimization scheme set is empty. If it is empty, the threshold down-regulation process is started, and the feature similarity screening threshold is reduced by 0.05 fixed step by step. Each time the threshold is reduced, the historical optimization schemes of the scheme library are retrieved again, the cosine similarity algorithm is used to calculate the matching degree, and the feature consistency check is carried out at the same time, until at least 3 effective candidate schemes are screened out. The threshold bottom line is set to not less than 0.6 during the down-regulation process, to avoid excessive threshold reduction leading to a significant decrease in scheme adaptability. At the same time, the schemes with relatively higher similarity are preferentially retained after each round of down-regulation, taking into account the number requirement and basic adaptability. If it still fails to meet the requirement after being reduced to the bottom line, the historical schemes of similar operating conditions are supplemented and the consistency check standard is fine-tuned to ensure the sufficiency of the number of candidate schemes. This mechanism not only guarantees that the scheme screening has sufficient sample support, but also strictly controls the adaptation bottom line, laying a solid foundation for subsequent priority sorting and optimal scheme selection, and avoiding the stagnation of the power flow optimization process due to the lack of candidate schemes.

[0109] S434, the feature consistency check is carried out on the schemes in the candidate optimization scheme set, and the schemes with an adaptation degree to the current power grid operating condition parameter lower than 80% are excluded, and the effective candidate schemes are retained.

[0110] Specifically, the historical working condition core parameters corresponding to each candidate scheme and the current power grid real-time working condition core dimensions (voltage, frequency, load rate) are extracted first, and the adaptation degree is calculated according to the preset weight. The key optimization direction of the scheme is checked whether it is consistent with the current power flow regulation demand. The candidate schemes are checked one by one. If the adaptation degree is less than 80%, it means that the scheme is not matched with the current working condition, which is easy to cause optimization deviation, and is directly excluded. If the adaptation degree meets the standard, it is retained as an effective candidate scheme. Both feature similarity and working condition adaptation are considered in the verification to avoid the limitation of single dimension screening. The adaptation details and core advantages of the effective scheme are recorded to provide support for subsequent priority sorting, ensuring that the retained scheme not only fits the current power grid operating state, but also has practical optimization effect, improving the reliability of the final power flow scheme.

[0111] S44, sort the candidate optimization scheme set based on the candidate scheme priority sorting rule, and select the highest ranked candidate optimization scheme as the intelligent power grid dynamic power flow optimization scheme.

[0112] Specifically, the feature similarity is used as the primary sorting basis first, and the schemes with higher similarity are ranked in the front row, and the schemes with higher adaptation degree to the current working condition are preferentially locked. For the schemes with similar similarity, the power loss reduction amplitude corresponding to each scheme is compared, and the scheme with larger loss reduction amplitude is preferentially selected to meet the power grid energy-saving operation demand. Finally, for the schemes with equivalent indicators in the previous two items, the historical debugging cost is calculated, and the scheme with lower cost is preferentially selected to consider the optimization landing economy. The feature consistency of the scheme and the current working condition is checked during the sorting process to ensure that the sorting does not deviate from the adaptation core. After completing the full-level sorting, the highest ranked candidate optimization scheme is selected as the final intelligent power grid dynamic power flow optimization scheme, which not only ensures the accurate adaptation of the scheme, but also realizes the minimization of loss and optimization of cost, improving the practicality and efficiency of power flow regulation.

[0113] S5, combine the intelligent power grid dynamic power flow optimization scheme with the deviation compensation logic of the historical power flow control parameter to perform staged debugging on the intelligent power grid power flow control device, and collect the power flow change value after debugging in real time.

[0114] In the embodiment of the application, the intelligent power grid dynamic power flow optimization scheme is combined with the deviation compensation logic of the historical power flow control parameter to perform staged debugging on the intelligent power grid power flow control device, and the power flow change value after debugging is collected in real time, which includes the following steps:

[0115] S51, retrieve the device debugging deviation record, control response delay data and past optimization scheme adaptation effect in the historical power flow control parameter, and set the device debugging benchmark parameter range in combination with the intelligent power grid dynamic power flow optimization scheme;

[0116] Specifically, the historical full-amount data of power grid tide flow control is comprehensively retrieved, and three types of core information, including equipment debugging deviation record, control response delay data and adaptation effect of previous optimization scheme, are extracted. After classification and analysis, the common deviation interval of equipment debugging, the reasonable threshold of response delay and the key influencing factors of scheme adaptation effect are determined. Then, combined with the core requirements of the selected smart grid dynamic tide flow optimization scheme, the equipment debugging reference parameter range is set based on historical data. The reference parameter range not only avoids the deviation exceeding problem in previous debugging, but also matches the requirements of control response speed of the scheme, and the boundary of the reference range is calibrated by referring to the historical parameter interval with good adaptation effect. After setting, the reference parameter range is verified to ensure that it meets the equipment safety operation specification and is highly consistent with the dynamic tide flow optimization goal, providing clear standards for subsequent field debugging, reducing debugging deviation and response delay, ensuring the accurate landing of the optimization scheme, and improving the stability and adaptability of tide flow control.

[0117] S52, pre-debugging, fine-tuning and stable verification of tide flow control equipment based on equipment debugging reference parameter range are carried out in three stages until the equipment running state matches the smart grid dynamic tide flow optimization scheme;

[0118] Specifically, based on the preset equipment debugging reference parameter range, the tide flow control equipment is debugged in three stages until the equipment running state matches the smart grid dynamic tide flow optimization scheme. In the first stage, pre-debugging is carried out, the initial running parameters of the equipment are set according to the reference parameter range, the basic debugging such as wiring and starting and stopping is completed, the equipment faults are checked out, it is ensured that the equipment can normally respond to the control command, and the invalid running state is eliminated. In the second stage, fine-tuning is promoted, the core control parameters are adjusted one by one according to the upper and lower limits of the reference parameters, the deviation is corrected combined with the requirements of the tide flow optimization scheme, the parameter fluctuation range is narrowed, and the equipment running precision is ensured to meet the requirements of the scheme. In the third stage, stable verification is implemented, the equipment running data is continuously monitored, the parameter stability and response timeliness are verified, and it is confirmed that there is no deviation exceeding and response delay problem. The equipment state and the adaptation degree of the optimization scheme are compared throughout the process, if not up to standard, the fine-tuning and verification are repeated, if up to standard, the debugging is terminated, it is ensured that the equipment stably supports the landing of the tide flow optimization scheme, and the reliability of power grid operation is improved.

[0119] S53, when the equipment is stably running, the tide flow change value of the power grid is continuously collected and stored according to time nodes.

[0120] Specifically, when the power flow control device completes the three-stage debugging and realizes stable operation, the power grid power flow data continuous collection mechanism is immediately started, and key data such as power loss, power flow overrun related parameters, voltage frequency fluctuation and core node power flow change value are accurately collected, ensuring that the collection frequency is consistent with the 24-hour time node of the previously constructed feature matrix, and ensuring the time sequence coherence of the data. During the collection process, real-time verification is carried out simultaneously, and abnormal and distorted data are eliminated to avoid invalid information retention. Subsequently, the data are strictly stored according to the preset time node, and the corresponding period power flow change full data are sorted and collected in units of hours, while binding the power grid working condition label, device operation parameter and optimization scheme execution situation of the day, and establishing a structured data account. When storing, a special database is built and hierarchical indexing is set to facilitate subsequent quick retrieval and analysis, and to provide complete and accurate time sequence data support for subsequent scheme adaptation effect review, dynamic weighting rule optimization and iterative optimization, and to lay a data foundation for power grid power flow closed-loop regulation.

[0121] S6, a power flow change safety threshold is preset, the power grid power flow change value is safety verified based on the power flow change safety threshold, and a closed loop optimization of the intelligent power grid dynamic power flow optimization scheme is formed according to the safety verification result.

[0122] Specifically, the power grid power flow change safety threshold is preset in combination with the power grid safety operation standard and the historical stable working condition data, the safety range of the core indicators such as power loss, power flow overrun probability and voltage frequency fluctuation is determined, and the threshold red line is drawn to ensure that the power grid operation specification is not broken. After the device is stably operated, the continuously collected power flow change value is extracted, and the safety verification is carried out item by item by comparing with the preset safety threshold, whether the indicators are in the safety range is judged, and the deviation data and the over-standard nodes are recorded synchronously. If the verification is passed, the current intelligent power grid dynamic power flow optimization scheme is maintained for execution; if the indicators are over-standard or close to the threshold, the dynamic adjustment mechanism is immediately started, and the weighting rule, device parameter or regulation strategy in the scheme is optimized. After adjustment, the data are collected and verified again, and the closed loop optimization system of "collection-verification-adjustment-landing" is formed, the real-time working condition change of the power grid is continuously adapted, and the power grid power flow is always stably in the safety range, so that the accuracy and reliability of dynamic regulation are improved.

[0123] The above only describes the preferred embodiments of the present application and is not intended to limit the present application. Any modification, equivalent replacement, improvement, etc. made within the spirit and principles of the present application shall be included in the protection scope of the present application.

Claims

1. A smart grid adaptive closed-loop power flow optimization control method, characterized in that, Includes the following steps: S1. Acquire multi-source core data, real-time power grid operating parameters, and historical power flow control parameters from the power generation side, transmission side, and user side of the smart grid, and establish a smart grid dynamic power flow optimization scheme library, wherein the historical optimization schemes in the smart grid dynamic power flow optimization scheme library are associated with the feature matrix in a one-to-one correspondence. S2. Extract the power flow fluctuation feature parameters from the multi-source core data, preset the multi-source data fusion association rules, and map the power flow fluctuation feature parameters into a power grid power flow feature mapping matrix according to the multi-source data fusion association rules. S3. Based on the real-time operating parameters of the power grid, set multi-dimensional dynamic weighting rules, construct an optimization model using the BP neural network algorithm, and iteratively optimize and adjust the power flow feature mapping matrix of the power grid with the dual optimization objectives of minimizing power grid power loss and reducing the probability of power flow exceeding limits, to obtain the optimized power flow feature mapping matrix. S4. Preset dynamic power flow scheme matching rules, and perform feature similarity matching between the optimized power flow feature mapping matrix and the smart grid dynamic power flow optimization scheme library based on the dynamic power flow scheme matching rules, and filter and sort to obtain smart grid dynamic power flow optimization schemes. S5. Combine the aforementioned smart grid dynamic power flow optimization scheme with the deviation compensation logic of historical power flow control parameters, perform phased debugging of the smart grid power flow control equipment, and collect the power flow change value of the grid after debugging in real time. S6. Preset a power flow change safety threshold, perform safety verification on the power flow change value of the power grid based on the power flow change safety threshold, and dynamically adjust the smart grid dynamic power flow optimization scheme according to the safety verification result to form a closed-loop optimization.

2. The adaptive closed-loop power flow optimization control method for smart grids as described in claim 1, characterized in that, In step S1, acquiring multi-source core data includes: collecting output fluctuation rate and power stability coefficient from the power generation side through the new energy monitoring system and unit monitoring platform; collecting line load rate and voltage deviation rate from the transmission side through the SCADA system and WAMS system; and collecting load mutation rate and electricity consumption period distribution coefficient from the user side through smart meters and load monitoring terminals.

3. The adaptive closed-loop power flow optimization control method for smart grids as described in claim 1, characterized in that, In step S2, the power flow fluctuation characteristic parameters include output fluctuation rate, power stability coefficient, line load rate, voltage deviation rate, load mutation rate, and electricity consumption period distribution coefficient.

4. The adaptive closed-loop power flow optimization control method for smart grids as described in claim 1, characterized in that, In step S2, the preset multi-source data fusion association rules include: setting the weight of power generation side data to 40%, the weight of transmission side data to 30%, and the weight of user side data to 30%, and configuring initial weights for the power flow fluctuation characteristic parameters of each side, wherein the power generation side output fluctuation rate weight is 0.25, the power stability coefficient weight is 0.15, the transmission side line load rate weight is 0.2, the voltage deviation rate weight is 0.1, the user side load mutation rate weight is 0.15, the electricity consumption period distribution coefficient weight is 0.15, and the sum of the initial weights of all characteristic parameters is 1.

5. The adaptive closed-loop power flow optimization control method for smart grids as described in claim 1, characterized in that, In step S3, the optimization model constructed using the BP neural network algorithm includes: constructing a three-layer structure of input layer-hidden layer-output layer, wherein the number of nodes in the input layer is the total number of elements in the power flow characteristic mapping matrix, the number of nodes in the output layer is 2 and corresponds to the predicted power loss value and the predicted power flow over-limit probability value, respectively, and the number of nodes in the hidden layer is determined by empirical formula based on the number of nodes in the input layer and the number of nodes in the output layer.

6. The adaptive closed-loop power flow optimization control method for smart grids as described in claim 5, characterized in that, The empirical formula is: ； Where n is the number of hidden layer nodes, m is the number of input layer nodes, k is the number of output layer nodes, and a is an adjustment constant between 10 and 20.

7. The adaptive closed-loop power flow optimization control method for smart grids as described in claim 1, characterized in that, In step S4, the dynamic power flow scheme matching rule is a cosine similarity algorithm, and the formula for the cosine similarity algorithm is: ， Where A is the optimized power flow feature mapping matrix, and B is the feature matrix of historical optimized schemes in the scheme library.

8. The adaptive closed-loop power flow optimization control method for smart grids as described in claim 7, characterized in that, The preset feature similarity screening threshold is 0.

85. When the calculated feature similarity is greater than or equal to 0.85, the corresponding historical optimization scheme is included in the candidate optimization scheme set.

9. The adaptive closed-loop power flow optimization control method for smart grids as described in claim 1, characterized in that, In step S5, the phased commissioning of the smart grid power flow control equipment includes: pre-commissioning stage, fine-tuning stage, and stability verification stage.

10. The adaptive closed-loop power flow optimization control method for smart grids as described in claim 1, characterized in that, In step S6, the preset power flow change safety threshold is set based on the power grid safe operation standard and historical stable operating condition data. The power flow change safety threshold includes the power loss safety range, the power flow over-limit probability safety range, and the voltage frequency fluctuation safety range.

Images