Intelligent control method for flexible interconnection of power distribution area

Abstract

The invention discloses an intelligent control method for flexible interconnection of a power distribution area, and relates to the technical field of power distribution network control, and the method comprises the steps: S1, carrying out the data collection of a target power distribution area, obtaining an initial data set, and obtaining a static topological parameter based on the operation historical data processing; s2, processing based on the operation historical data to obtain a dynamic constraint envelope; s3, establishing a system differential flat state equation based on the static topological parameters and the dynamic constraint envelope, and performing security domain interception based on a flat output space to obtain a feasible solution hypersurface; s4, performing path search on the feasible solution hypersurface through an improved ant colony optimization algorithm, updating pheromone increment and completing iterative calculation; and S5, selecting a global optimal path based on an iterative calculation result, and generating an SOP control instruction of the target power distribution area. The defects that only a local optimal solution is concerned and response lag is easy to occur when a station area control instruction is generated in the prior art are overcome.

Description

Technical Field

[0001] The present invention relates to the technical field of distribution network control, and in particular to an intelligent control method for flexible interconnection of distribution network areas. Background Art

[0002] In recent years, thanks to the rapid development of power electronics, intelligent control, and computer information technology, new loads such as distributed photovoltaics and electric vehicles have begun to be connected to distribution networks on a large scale. The closed-loop design and open-loop operation of traditional distribution networks are no longer able to meet the diverse load demands of power supply reliability and renewable energy consumption. Flexible interconnection of distribution substations offers power electronics conversion and intelligent control capabilities, and aligns with the trend toward green and intelligent development. Therefore, developing flexible interconnection control methods for distribution substations with real-time control capabilities to address issues such as inter-station load imbalance and node voltage and current overshooting is a promising application direction.

[0003] At present, the Chinese invention patent application with application number CN202510254649.4 discloses a method and system for intelligent control of photovoltaic energy storage in distribution substations. The application includes: collecting data from photovoltaic inverters, energy storage devices and grid equipment, and through real-time monitoring of the operating status, current, voltage, active power, reactive power and power factor of the aforementioned equipment, constructing a voltage over-limit control strategy for the grid-connected nodes of the photovoltaic inverters in the distribution substation based on data analysis to control the photovoltaic inverters. The present invention solves the problem of voltage over-limit at the nodes where the photovoltaic inverters in the substation are connected to the grid, and increases the amount of power generated on the grid. However, the application only relies on real-time operating data to realize the control of the photovoltaic inverter, and lacks consideration of the inherent physical characteristics of the substation such as the node impedance matrix and historical operating laws such as power fluctuations. In addition, only the output power of the photovoltaic inverter is adjusted to solve the problem of abnormal voltage at local nodes, but the overall safe operation boundary of the substation is not systematically defined, and the adjustment range is limited. Summary of the Invention

[0004] The technical problem addressed by this invention is that existing technologies often only analyze and control local issues related to a single device, lacking overall control over the operation of the distribution substation, resulting in a significantly limited scope of application. Monitoring and control rely on real-time operational data, but insufficiently consider the inherent physical characteristics and historical operating patterns of the distribution substation. Furthermore, existing technologies employ relatively simple constraint settings, failing to address global voltage and current limit violations.

[0005] In order to solve the above technical problems, the present invention provides the following technical solutions: An intelligent control method for flexible interconnection of distribution substations; Step S1: collect data from the target distribution substation area to obtain an initial data set, and process the historical operation data to obtain static topology parameters; Step S2, obtaining a dynamic constraint envelope based on the operation history data processing; Step S3, establishing a system differential flat state equation based on static topological parameters and dynamic constraint envelope, and performing a safety domain interception based on the flat output space to obtain a feasible solution hypersurface; Step S4, performing path search on the feasible solution hypersurface using the improved ant colony optimization algorithm, updating the pheromone increment and completing the iterative calculation; Step S5: Select the global optimal path based on the iterative calculation result and generate the SOP control instruction of the target distribution station area.

[0006] Preferably, the initial data set includes network topology data of the target distribution substation area, system parameters of the flexible interconnection device, real-time operation data and operation history data; Step S1: Processing the operation history data in the initial data set to obtain static topology parameters, wherein the static topology parameters include a node impedance matrix and a SOP capacity limit. The processing logic includes: Based on the network topology data of the target distribution substation area, the resistance data and reactance data between each grid node are statistically obtained; Establish a node impedance matrix Z based on the resistance data and reactance data between each grid node; The calculation expression for establishing the node impedance matrix Z is: ; Where Z represents the node impedance matrix, represents the per-unit resistance between node m and node n, represents the per-unit reactance value between node m and node n, where m and n represent the grid node numbers and j represents an imaginary number; The grid nodes include key nodes and secondary nodes; The key nodes include the transformer low-voltage side busbar node and the main feeder head end node; The secondary nodes include branch box nodes and user meter box nodes; The real-time operation data includes the SOP remaining capacity, the real-time voltage of each grid node and the real-time current of each grid node; The SOP capacity limit is obtained based on the system parameters of the flexible interconnect device.

[0007] Preferably, in step S2, a dynamic constraint envelope is obtained by processing the operation history data; The dynamic constraint envelope includes the node active power injection interval, net injected active power and net injected reactive power; The operation history data includes the historical voltage of each grid node, the historical current of each grid node, the active power history data and the reactive power history data; The operation history data is obtained by synchronous sampling at intervals of 15 minutes; Sort the discrete sampling points in the active power history data and select the maximum per-unit value of the active power history data As the upper limit of the interval, select the minimum per unit value As the lower limit of the interval, the node active power injection interval is obtained ; The expected value of active load is obtained by performing arithmetic mean calculation on all discrete sampling point values ​​in the active power history data. ; The reactive power historical data is fitted by the kernel density estimation algorithm to obtain the load power fluctuation probability density function, and the load power fluctuation probability density function is integrated to obtain the expected reactive load power value. ; Based on the power balance relationship of the target distribution station area and the KCL law, the power balance calculation is performed on the expected value of active load and the expected power value of reactive load to obtain the net injected active power and net injected reactive power. The calculation expressions are: ; ; in, represents the per-unit value of the net injected active power, Indicates the per-unit value of active power injected by the upper power grid. Indicates the expected power per unit of active load, represents the per-unit value of the net injected reactive power, Indicates the per-unit value of reactive power injected by the upper power grid. Indicates the expected power per unit of reactive load.

[0008] Preferably, in step S301, a system differential flat state equation is established based on static topology parameters and dynamic constraint envelope, and a flat output space is obtained by mapping. The processing logic includes: Define the voltage and current of the key node as flat output y, and the expression of flat output y is: ; in, represents the per-unit value of the key node voltage, represents the per-unit value of the key node current, T represents the matrix transpose, k and l represent the key node numbers; Node active power injection interval As a deterministic boundary, the flat output space is obtained by flat space mapping the system differential flat state equation based on the Lie derivative algorithm. Its calculation expression is: ; in, represents the voltage trend derivative term, represents the current trend derivative term, represents the voltage state evolution function, represents the current state evolution function, Indicates the SOP capacity limit per unit value.

[0009] Preferably, in step S302, a safe domain interception is performed based on the flat output space to obtain a feasible solution hypersurface, and its processing logic includes: Perform an inverse operation on the impedance matrix Z and establish a safety scaling diagonal matrix based on the rated voltage and rated current of the target distribution station area. Multiply the impedance matrix and the safety scaling diagonal matrix to obtain the safety constraint matrix P, which is calculated as follows: ; Where P represents the safety constraint matrix, Indicates the rated voltage per unit value, Indicates the per-unit value of rated current. represents the inverse matrix of the impedance matrix; The safety constraint matrix P is used as the weight matrix of the Lyapunov function, and the flat output space is intercepted by the ellipsoidal safety region to obtain the feasible solution hypersurface, whose calculation expression is: ; in, represents the feasible solution hypersurface.

[0010] Preferably, in step S401, generating an initial pheromone distribution based on a feasible solution hypersurface, the processing logic includes: Perform uniform grid division on the feasible solution hypersurface, divide the feasible solution space into equally spaced grid points based on the preset voltage dimension step size and current dimension step size, number the surface space grid points and arrange ants to obtain the ant initial position set; The initial pheromone concentration was calculated by taking the inverse of the SOP capacity limit per unit value.

[0011] Preferably, in step S402 , the real-time voltage and real-time current of each key node are substituted into the flat output space, a real-time current trend derivative term is calculated, and the discretization process is performed to obtain the real-time node current gradient; The real-time node voltage deviation is obtained by calculating the difference between the rated voltage and the real-time voltage of each key node; The heuristic factors are calculated for the real-time node current gradient and the real-time node voltage deviation respectively, and the calculation expressions are as follows: ; in, represents the heuristic factor, Indicates the real-time node voltage deviation per unit value, It represents the per-unit value of the real-time node current gradient, and s and t represent the numbers of the surface space grid points.

[0012] Preferably, in step S403, a path search is performed on the feasible solution hypersurface to obtain an elite ant path and calculate the pheromone increment. The processing logic includes: The pheromone concentration and heuristic factor are calculated based on the ant optimization algorithm to obtain the path transfer probability distribution; Generate a grid point sequence based on the path transition probability distribution to obtain the complete ant path set; Calculate the path loss value corresponding to each path in the ant complete path set respectively, and the calculation expression is: ; represents the path loss value corresponding to the a-th ant, u represents the step number of the grid point sequence corresponding to the a-th ant, and U represents the total number of steps of the grid point sequence corresponding to the a-th ant; Sort all path loss values ​​from small to large, select the minimum value to get the optimal loss value, and select the grid point sequence corresponding to the optimal loss value from the ant complete path set as the elite ant path; By introducing the Fibonacci ratio pheromone update rule to calculate the elite ant path, the pheromone increment is obtained, and its calculation expression is: ; in, represents the pheromone increment, represents the current optimal loss value, H represents the total amount of pheromone released, k represents the number of iterations, Indicates the preset maximum number of iterations, Represents the Fibonacci ratio.

[0013] Preferably, in step S404, the pheromone increment and the initial pheromone are added to obtain a first updated pheromone, and the first updated pheromone is substituted back into step S403 for iterative calculation and determination of an iteration termination condition; When the iteration termination condition is reached, the iteration is stopped and the iterative calculation result is obtained; The iterative calculation results include the optimal loss value and elite ant path corresponding to the last iterative calculation; The iteration termination condition includes reaching a preset maximum number of iterations or the optimal solution being continuously stable; The continuous stability of the optimal solution includes that the difference between the optimal loss value of the most recent five iterations and the historical optimal loss value is less than 0.01.

[0014] Preferably, in step S5, based on the iterative calculation result, the elite ant path corresponding to the last iterative calculation is selected to obtain the global optimal path; Select the end point coordinates of the grid point sequence corresponding to the global optimal path to obtain the global optimal flat output; The global optimal flat output includes the real-time optimal voltage setting value and the real-time optimal current setting value; The optimal voltage setting value is mapped to the active power that the SOP should output, and the real-time optimal current setting value is mapped to obtain the reactive power that the SOP should output; The SOP control instruction of the target distribution station area is obtained based on the active power that the SOP should output and the reactive power that the SOP should output.

[0015] Beneficial effects of the present invention: This application takes into account the static physical characteristics and dynamic operating laws of the distribution substation at the same time during regulation, performs static topology parameter extraction and dynamic constraint envelope calculation, and avoids distortion through physical constraints while retaining the historical characteristics of load fluctuations. This application establishes a differential flat model to reduce the complexity of the power grid system state, and obtains a feasible solution hypersurface for the high-dimensional ellipsoid interception, clarifying the global safety domain of the substation. This application fully considers the complex characteristics of the distribution substation data, improves the ant colony optimization algorithm in a targeted manner, overcomes the problems of falling into local optimality and slow convergence, and effectively solves the defects of only local optimality and response lag when generating distribution substation control instructions. This application realizes the intelligent control of the substation as a whole, and at the same time takes into account data processing accuracy and computational efficiency in the process of flexible interconnection regulation of distribution substations, which is conducive to improving the safety, economy and robustness of system control. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Figure 1 A schematic diagram of the basic flow of an intelligent control method for flexible interconnection of distribution substations provided by one embodiment of the present invention; Figure 2 A schematic diagram of a dynamic constraint envelope generation process according to an embodiment of the present invention; Figure 3 A schematic diagram of the iterative calculation process of the improved ant colony optimization provided by one embodiment of the present invention. DETAILED DESCRIPTION

[0017] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the specific implementation methods of the present invention are described in detail below in conjunction with the drawings. It is obvious that the described embodiments are only part of the embodiments of the present invention, but not all of the embodiments.

[0018] Reference Figure 1-3 , as one embodiment of the present invention, provides an intelligent control method for flexible interconnection of distribution substations, comprising: Step S1: collect data from the target distribution substation area to obtain an initial data set, and process the historical operation data to obtain static topology parameters; Step S2, obtaining a dynamic constraint envelope based on the operation history data processing; Step S3, establishing a system differential flat state equation based on static topological parameters and dynamic constraint envelope, and performing a safety domain interception based on the flat output space to obtain a feasible solution hypersurface; Step S4, performing path search on the feasible solution hypersurface using the improved ant colony optimization algorithm, updating the pheromone increment and completing the iterative calculation; Step S5: Select the global optimal path based on the iterative calculation result and generate the SOP control instruction of the target distribution station area.

[0019] In this embodiment, the initial data set includes network topology data of the target distribution substation area, system parameters of the flexible interconnection device, real-time operation data and operation history data; Step S1: Processing the operation history data in the initial data set to obtain static topology parameters, wherein the static topology parameters include a node impedance matrix and a SOP capacity limit. The processing logic includes: Based on the network topology data of the target distribution substation area, the resistance data and reactance data between each grid node are statistically obtained; Establish a node impedance matrix Z based on the resistance data and reactance data between each grid node; The calculation expression for establishing the node impedance matrix Z is: ; Where Z represents the node impedance matrix, represents the per-unit resistance between node m and node n, represents the per-unit reactance value between node m and node n, where m and n represent the grid node numbers and j represents an imaginary number; The grid nodes include key nodes and secondary nodes; The key nodes include the transformer low-voltage side busbar node and the main feeder head end node; The secondary nodes include branch box nodes and user meter box nodes; The real-time operation data includes the SOP remaining capacity, the real-time voltage of each grid node and the real-time current of each grid node; The SOP capacity limit is obtained based on the system parameters of the flexible interconnect device.

[0020] This application uses a per-unit value system with a reference voltage of 0.4kV and a reference power of 10MVA. All power, current, and impedance parameters involved have been dimensionlessly processed according to this reference value. This helps ensure the accuracy and stability of the matrix calculations in this application. In this embodiment, step S2 is to obtain a dynamic constraint envelope based on the operation history data; The dynamic constraint envelope includes the node active power injection interval, net injected active power and net injected reactive power; The operation history data includes the historical voltage of each grid node, the historical current of each grid node, the active power history data and the reactive power history data; The operation history data is obtained by synchronous sampling at intervals of 15 minutes; Sort the discrete sampling points in the active power history data and select the maximum per-unit value of the active power history data As the upper limit of the interval, select the minimum per unit value As the lower limit of the interval, the node active power injection interval is obtained ; The expected value of active load is obtained by performing arithmetic mean calculation on all discrete sampling point values ​​in the active power history data. ; The reactive power historical data is fitted by the kernel density estimation algorithm to obtain the load power fluctuation probability density function, and the load power fluctuation probability density function is integrated to obtain the expected reactive load power value. ; Based on the power balance relationship of the target distribution station area and the KCL law, the power balance calculation is performed on the expected value of active load and the expected power value of reactive load to obtain the net injected active power and net injected reactive power. The calculation expressions are: ; ; in, represents the per-unit value of the net injected active power, Indicates the per-unit value of active power injected by the upper power grid. Indicates the expected power per unit of active load, represents the per-unit value of the net injected reactive power, Indicates the per-unit value of reactive power injected by the upper power grid. Indicates the expected power per unit value of reactive load; The processing logic for fitting reactive power historical data using the kernel density estimation algorithm includes: The Gaussian kernel function is used to perform weighted superposition processing on the discrete sample points in the reactive power historical data to obtain a continuous probability density curve. The kernel window width parameter of the Gaussian kernel function is optimized based on the Silverman criterion to balance the fitting accuracy and smoothness. At the same time, the power cannot be negative as the physical constraint condition of reactive power. The boundary reflection method is used to correct the probability density distortion at the distribution edge of the continuous probability density curve to obtain the load reactive power probability density function.

[0021] In this embodiment, step S301 establishes a system differential flat state equation based on static topology parameters and dynamic constraint envelope, and maps to obtain a flat output space. The processing logic includes: Define the voltage and current of the key node as flat output y, and the expression of flat output y is: ; in, represents the per-unit value of the key node voltage, represents the per-unit value of the key node current, T represents the matrix transpose, k and l represent the key node numbers; Node active power injection interval As a deterministic boundary, the flat output space is obtained by flat space mapping the system differential flat state equation based on the Lie derivative algorithm. Its calculation expression is: ; in, represents the voltage trend derivative term, represents the current trend derivative term, represents the voltage state evolution function, represents the current state evolution function, Indicates the SOP capacity limit per unit value.

[0022] Among them, the definition of flat output y is not affected by subsequent updates of the state equation and is the basic coordinate system of the safety domain.

[0023] In this embodiment, step S302 performs a safe domain interception based on the flat output space to obtain a feasible solution hypersurface, and its processing logic includes: Perform an inverse operation on the impedance matrix Z and establish a safety scaling diagonal matrix based on the rated voltage and rated current of the target distribution station area. Multiply the impedance matrix and the safety scaling diagonal matrix to obtain the safety constraint matrix P, which is calculated as follows: ; Where P represents the safety constraint matrix, Indicates the rated voltage per unit value, Indicates the per-unit value of rated current. represents the inverse matrix of the impedance matrix; The safety constraint matrix P is used as the weight matrix of the Lyapunov function, and the flat output space is intercepted by the ellipsoidal safety region to obtain the feasible solution hypersurface, whose calculation expression is: ; in, represents the feasible solution hypersurface.

[0024] The rated voltage coefficient of 0.05 in the safety constraint matrix P represents the voltage tolerance of ±5%, and the rated current coefficient of 1.1 represents the line thermal stability limit. The safety constraint matrix P satisfies the positive definiteness verification, and its minimum eigenvalue is greater than zero; The geometric form of the feasible solution hypersurface is a high-dimensional ellipsoid, which describes all SOP operating points that meet the N-1 safety criterion in the power grid field. The N-1 safety criterion means that a single line can still maintain safe operation after a fault.

[0025] Steps S2 and S3 are calculated based on the target distribution substation's inherent parameters and historical data, extracting its operating characteristics. The voltages and currents involved in steps S2 and S3 are sampled from historical data. Starting in step S4, real-time data is incorporated into the calculations to achieve intelligent control of the flexible interconnection of distribution substations. The voltages and currents in steps S4 and beyond are real-time measured data.

[0026] In this embodiment, step S401 generates an initial pheromone distribution based on a feasible solution hypersurface. The processing logic includes: Perform uniform grid division on the feasible solution hypersurface, divide the feasible solution space into equally spaced grid points based on the preset voltage dimension step size and current dimension step size, number the surface space grid points and arrange ants to obtain the ant initial position set; The initial pheromone concentration was calculated by taking the inverse of the SOP capacity limit per unit value.

[0027] In this embodiment, in step S402 , the real-time voltage and real-time current of each key node are substituted into the flat output space, the real-time current trend derivative term is calculated, and the discretization process is performed to obtain the real-time node current gradient; The real-time node voltage deviation is obtained by calculating the difference between the rated voltage and the real-time voltage of each key node; The heuristic factors are calculated for the real-time node current gradient and the real-time node voltage deviation respectively, and the calculation expressions are as follows: ; in, represents the heuristic factor, Indicates the real-time node voltage deviation per unit value, It represents the per-unit value of the real-time node current gradient, and s and t represent the numbers of the surface space grid points.

[0028] In this embodiment, step S403 performs path search on the feasible solution hypersurface to obtain the elite ant path and calculates the pheromone increment. The processing logic includes: The pheromone concentration and heuristic factor are calculated based on the ant optimization algorithm to obtain the path transfer probability distribution; Generate a grid point sequence based on the path transition probability distribution to obtain the complete ant path set; Calculate the path loss value corresponding to each path in the ant complete path set respectively, and the calculation expression is: ; represents the path loss value corresponding to the a-th ant, u represents the step number of the grid point sequence corresponding to the a-th ant, and U represents the total number of steps of the grid point sequence corresponding to the a-th ant; Sort all path loss values ​​from small to large, select the minimum value to get the optimal loss value, and select the grid point sequence corresponding to the optimal loss value from the ant complete path set as the elite ant path; By introducing the Fibonacci ratio pheromone update rule to calculate the elite ant path, the pheromone increment is obtained, and its calculation expression is: ; in, represents the pheromone increment, represents the current optimal loss value, H represents the total amount of pheromone released, k represents the number of iterations, Indicates the preset maximum number of iterations, Represents the Fibonacci ratio.

[0029] Among them, in order to adapt to the high fluctuation source and load scenario in the distribution area while taking into account the calculation efficiency, the preset maximum number of iterations is set to 120, and the Fibonacci ratio Take 0.618, and the total pheromone release H is set to 0.9 in residential areas and 2.0 in industrial areas.

[0030] In this embodiment, in step S404, the pheromone increment and the initial pheromone are added to obtain a first updated pheromone, and the first updated pheromone is substituted back into step S403 for iterative calculation and determination of the iterative termination condition. When the iteration termination condition is reached, the iteration is stopped and the iterative calculation result is obtained; The iterative calculation results include the optimal loss value and elite ant path corresponding to the last iterative calculation; The iteration termination condition includes reaching a preset maximum number of iterations or the optimal solution being continuously stable; The continuous stability of the optimal solution includes that the difference between the optimal loss value of the most recent five iterations and the historical optimal loss value is less than 0.01.

[0031] Among them, the initial pheromone concentration is set based on the inverse of the SOP capacity limit, which prioritizes guiding ants to explore high-capacity areas and helps avoid invalid searches; the heuristic factor is calculated by combining real-time voltage deviation and current gradient to enhance the sensitivity to the deviation of the current operating state from the target; only the elite ant path, that is, the path corresponding to the optimal loss value, is incrementally updated with pheromone to accelerate the accumulation of pheromones in the dominant path, and the Fibonacci proportional coefficient is introduced for iterative calculation to improve the calculation accuracy of the optimal solution.

[0032] In this embodiment, in step S5, based on the iterative calculation results, the elite ant path corresponding to the last iterative calculation is selected to obtain the global optimal path; Select the end point coordinates of the grid point sequence corresponding to the global optimal path to obtain the global optimal flat output; The global optimal flat output includes the real-time optimal voltage setting value and the real-time optimal current setting value; The optimal voltage setting value is mapped to the active power that the SOP should output, and the real-time optimal current setting value is mapped to obtain the reactive power that the SOP should output. Among them, mapping the optimal voltage setting value to SOP active power and mapping the real-time optimal current setting value to SOP reactive power are achieved based on the voltage-active power corresponding parameter table and current-reactive power corresponding parameter table pre-calibrated by the flexible interconnection device SOP of the target distribution station area.

[0033] Those skilled in the art will appreciate that embodiments of the present invention may provide methods, systems, or computer program products. Therefore, the present invention may take the form of an entirely hardware embodiment, an entirely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media containing computer-usable program code. The storage medium may be implemented by any type of volatile or non-volatile storage device, or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic storage, flash memory, magnetic disk, or optical disk. These computer program instructions may also be stored in a computer readable memory that can direct a computer or other programmable data processing device to work in a specific manner, so that the instructions stored in the computer readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 a process or multiple processes and / or boxes Figure 1 The function specified in one or more boxes.

[0034] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit the present invention. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present invention, which should all be included in the scope of the claims of the present invention.

Claims

1. An intelligent control method for flexible interconnection of distribution substations, characterized in that: include: Step S1: collect data from the target distribution substation area to obtain an initial data set, and process the historical operation data to obtain static topology parameters; Step S2, obtaining a dynamic constraint envelope based on the operation history data processing; Step S3, establishing the system differential flat state equation based on the static topological parameters and the dynamic constraint envelope, and performing a safety domain interception based on the flat output space to obtain a feasible solution hypersurface; Step S4, performing path search on the feasible solution hypersurface using the improved ant colony optimization algorithm, updating the pheromone increment and completing the iterative calculation; Step S5: Select the global optimal path based on the iterative calculation result and generate the SOP control instruction of the target distribution station area.

2. The intelligent control method for flexible interconnection of distribution substations according to claim 1, characterized in that: The initial data set includes network topology data of the target distribution substation area, system parameters of the flexible interconnection device, real-time operation data and operation history data; Step S1: Processing the operation history data in the initial data set to obtain static topology parameters, wherein the static topology parameters include a node impedance matrix and a SOP capacity limit. The processing logic includes: Based on the network topology data of the target distribution substation area, the resistance data and reactance data between each grid node are statistically obtained; Establish a node impedance matrix Z based on the resistance data and reactance data between each grid node; The calculation expression for establishing the node impedance matrix Z is: ； Where Z represents the node impedance matrix, represents the per-unit resistance between node m and node n, represents the per-unit reactance value between node m and node n, where m and n represent the grid node numbers and j represents an imaginary number; The grid nodes include key nodes and secondary nodes; The key nodes include the transformer low-voltage side busbar node and the main feeder head end node; The secondary nodes include branch box nodes and user meter box nodes; The real-time operation data includes the SOP remaining capacity, the real-time voltage of each grid node and the real-time current of each grid node; The SOP capacity limit is obtained based on the system parameters of the flexible interconnect device.

3. The intelligent control method for flexible interconnection of distribution substations according to claim 1, characterized in that: Step S2, processing the operation history data to obtain a dynamic constraint envelope; The dynamic constraint envelope includes the node active power injection interval, net injected active power and net injected reactive power; The operation history data includes the historical voltage of each grid node, the historical current of each grid node, the active power history data and the reactive power history data; The operation history data is obtained by synchronous sampling at intervals of 15 minutes; Sort the discrete sampling points in the active power history data and select the maximum per-unit value of the active power history data As the upper limit of the interval, select the minimum per unit value As the lower limit of the interval, the node active power injection interval is obtained ; The expected value of active load is obtained by performing arithmetic mean calculation on all discrete sampling point values ​​in the active power history data. ; The reactive power historical data is fitted by the kernel density estimation algorithm to obtain the load power fluctuation probability density function, and the load power fluctuation probability density function is integrated to obtain the expected reactive load power value. ; Based on the power balance relationship of the target distribution station area and the KCL law, the power balance calculation is performed on the expected value of active load and the expected power value of reactive load to obtain the net injected active power and net injected reactive power. The calculation expressions are: ； ； in, represents the per-unit value of the net injected active power, Indicates the per-unit value of active power injected by the upper power grid. Indicates the expected power per unit of active load, represents the per-unit value of the net injected reactive power, Indicates the per-unit value of reactive power injected by the upper power grid. Indicates the expected power per unit of reactive load.

4. The intelligent control method for flexible interconnection of distribution substations according to claim 1, characterized in that: Step S301: Establish a system differential flat state equation based on static topology parameters and dynamic constraint envelope, and map to obtain a flat output space. The processing logic includes: Define the voltage and current of the key node as flat output y, and the expression of flat output y is: ; in, represents the per-unit value of the key node voltage, represents the per-unit value of the key node current, T represents the matrix transpose, k and l represent the key node numbers; Node active power injection interval As a deterministic boundary, the flat output space is obtained by flat space mapping the system differential flat state equation based on the Lie derivative algorithm. Its calculation expression is: ; in, represents the voltage trend derivative term, represents the current trend derivative term, represents the voltage state evolution function, represents the current state evolution function, Indicates the SOP capacity limit per unit value.

5. The intelligent control method for flexible interconnection of distribution substations according to claim 1, characterized in that: Step S302: Performing a safety domain interception based on the flat output space to obtain a feasible solution hypersurface. The processing logic includes: Perform an inverse operation on the impedance matrix Z and establish a safety scaling diagonal matrix based on the rated voltage and rated current of the target distribution station area. Multiply the impedance matrix and the safety scaling diagonal matrix to obtain the safety constraint matrix P, which is calculated as follows: ； Where P represents the safety constraint matrix, Indicates the rated voltage per unit value, Indicates the per-unit value of rated current. represents the inverse matrix of the impedance matrix; The safety constraint matrix P is used as the weight matrix of the Lyapunov function, and the flat output space is intercepted by the ellipsoidal safety region to obtain the feasible solution hypersurface, whose calculation expression is: ； in, represents the feasible solution hypersurface.

6. The intelligent control method for flexible interconnection of distribution substations according to claim 1, characterized in that: Step S401: Generate an initial pheromone distribution based on a feasible solution hypersurface. The processing logic includes: Perform uniform grid division on the feasible solution hypersurface, divide the feasible solution space into equally spaced grid points based on the preset voltage dimension step size and current dimension step size, number the surface space grid points and arrange ants to obtain the ant initial position set; The initial pheromone concentration was calculated by taking the inverse of the SOP capacity limit per unit value.

7. The intelligent control method for flexible interconnection of distribution substations according to claim 1, characterized in that: Step S402 , substituting the real-time voltage and real-time current of each key node into the flat output space, calculating the real-time current trend derivative term, and performing discretization processing on the discretization to obtain the real-time node current gradient; The real-time node voltage deviation is obtained by calculating the difference between the rated voltage and the real-time voltage of each key node; The heuristic factors are calculated for the real-time node current gradient and the real-time node voltage deviation respectively, and the calculation expressions are as follows: ; in, represents the heuristic factor, Indicates the real-time node voltage deviation per unit value, It represents the per-unit value of the real-time node current gradient, and s and t represent the numbers of the surface space grid points.

8. The intelligent control method for flexible interconnection of distribution substations according to claim 1, characterized in that: Step S403: Path search is performed on the feasible solution hypersurface to obtain the elite ant path and calculate the pheromone increment. The processing logic includes: The pheromone concentration and heuristic factor are calculated based on the ant optimization algorithm to obtain the path transfer probability distribution; Generate a grid point sequence based on the path transition probability distribution to obtain the complete ant path set; Calculate the path loss value corresponding to each path in the ant complete path set respectively, and its calculation expression is: ； represents the path loss value corresponding to the a-th ant, u represents the step number of the grid point sequence corresponding to the a-th ant, and U represents the total number of steps of the grid point sequence corresponding to the a-th ant; Sort all path loss values ​​from small to large, select the minimum value to get the optimal loss value, and select the grid point sequence corresponding to the optimal loss value from the ant complete path set as the elite ant path; By introducing the Fibonacci ratio pheromone update rule to calculate the elite ant path, the pheromone increment is obtained, and its calculation expression is: ； in, represents the pheromone increment, represents the current optimal loss value, H represents the total amount of pheromone released, k represents the number of iterations, Indicates the preset maximum number of iterations, Represents the Fibonacci ratio.

9. The intelligent control method for flexible interconnection of distribution substations according to claim 1, characterized in that: Step S404, adding the pheromone increment to the initial pheromone to obtain a first updated pheromone, substituting the first updated pheromone back into step S403, iterating the calculation and determining the iteration termination condition; When the iteration termination condition is reached, the iteration is stopped and the iterative calculation result is obtained; The iterative calculation results include the optimal loss value and elite ant path corresponding to the last iterative calculation; The iteration termination condition includes reaching a preset maximum number of iterations or the optimal solution being continuously stable; The continuous stability of the optimal solution includes that the difference between the optimal loss value of the most recent five iterations and the historical optimal loss value is less than 0.

01.

10. The intelligent control method for flexible interconnection of distribution substations according to claim 1, characterized in that: Step S5: Based on the iterative calculation results, the elite ant path corresponding to the last iterative calculation is selected to obtain the global optimal path; Select the end point coordinates of the grid point sequence corresponding to the global optimal path to obtain the global optimal flat output; The global optimal flat output includes the real-time optimal voltage setting value and the real-time optimal current setting value; The optimal voltage setting value is mapped to the active power that the SOP should output, and the real-time optimal current setting value is mapped to obtain the reactive power that the SOP should output; The SOP control instruction of the target distribution station area is obtained based on the active power that the SOP should output and the reactive power that the SOP should output.

Images