Optimization decision-making method and system for transfer path of distribution network

Abstract

The invention relates to the technical field of power distribution network automation, in particular to an optimization decision-making method and system for a power distribution network transfer path. The method comprises the following steps: searching an optimal transfer path connected to a healthy power supply for a load to be transferred in a distribution network based on an improved ant colony algorithm; the improved ant colony algorithm comprises dynamic heuristic factors for guiding ants; the dynamic heuristic factor of any line is inversely proportional to the physical impedance and the fluctuation penalty factor of the line. According to the invention, the stability of the affected area after load transfer can be improved, and the occurrence of secondary faults is reduced.

Description

Technical Field

[0001] This application relates to the field of distribution network automation technology, and in particular to an optimization decision-making method and system for distribution network transfer paths. Background Technology

[0002] As a crucial component of the power system, the safe and stable operation of the distribution network is vital to social life and industrial production. When a line or equipment in the distribution network fails, or when planned maintenance is required, a transfer operation must be performed—that is, changing the switch combination—to quickly and safely transfer the load in the affected area to other healthy lines to ensure the continuity of power supply.

[0003] Currently, most decision-making methods for power distribution network transfer paths are based on graph theory search algorithms, such as Dijkstra's algorithm or Ant Colony Optimization (ACO). In traditional ant colony optimization applications, the algorithm finds the optimal path by simulating the behavior of ants searching for food. When ants choose the next node, they mainly rely on pheromone concentration and heuristic information. In the context of power distribution networks, this heuristic information is usually the reciprocal of the physical impedance in the line, that is, prioritizing the path with the shortest physical distance or the lowest resistance.

[0004] However, with the integration of high-penetration renewable energy sources such as distributed photovoltaic (PV) power, the operating environment of the power distribution network has undergone fundamental changes. The intermittency and volatility of PV power output have resulted in unprecedented dynamism and uncertainty in the magnitude and direction of power flow across lines. The physical impedance relied upon by traditional ant colony optimization algorithms can no longer reflect the real-time operating status and potential risks of the lines. For example, if a low-impedance path happens to be connected to a large-scale PV power plant with unstable output, a sudden weather change could cause a surge in PV output within minutes of a power transfer operation, leading to a severe overload and secondary faults after the load transfer is added. Therefore, existing technologies cannot effectively quantify and mitigate the dynamic overload risks caused by the volatility of renewable energy sources, resulting in a lack of robustness in power transfer decisions. Summary of the Invention

[0005] To improve the robustness of power transfer decisions and reduce overload situations after power transfer, this application provides an optimization decision-making method and system for power transfer paths in distribution networks.

[0006] Firstly, this application provides an optimization decision-making method for distribution network transfer paths, employing the following technical solution: An optimization decision-making method for distribution network transfer paths includes: finding the optimal transfer path for loads to be transferred in the distribution network to connect to a healthy power source based on an improved ant colony algorithm; The improved ant colony algorithm includes a dynamic heuristic factor to guide the ants; for any path, the dynamic heuristic factor is inversely proportional to the path's physical impedance and fluctuation penalty factor. The steps for obtaining the fluctuation penalty factor include: calculating the average output of the historical output data of the distributed power sources involved in the line; taking the sum of the line's base load and the load to be transferred in the area to be transferred as the total load; taking the ratio of the difference between the total load and the average output of the distributed power sources to the line's rated capacity as the expected net load rate; obtaining the relative fluctuation standard deviation that characterizes the degree of fluctuation in historical output data; constructing the fluctuation penalty factor based on the expected net load rate and the relative fluctuation standard deviation; and the fluctuation penalty factor being positively correlated with the expected net load rate and the relative fluctuation standard deviation.

[0007] The fluctuation penalty factor comprehensively considers the expected net load factor of the line after power transfer and the historical output fluctuation of distributed power sources in the line, which can quantify the risk of dynamic overload of the line caused by the fluctuation of new energy sources. As a result, the path decision in the ant colony algorithm is no longer limited to finding the path with the shortest physical distance or the lowest physical impedance, but can actively avoid those paths with low impedance but connected to unstable power sources or with excessively high load factors. This significantly improves the robustness and safety of power transfer decisions in distribution networks with a high proportion of new energy sources, and effectively avoids secondary faults caused by power flow fluctuations after power transfer.

[0008] Optionally, the steps for constructing a volatility penalty factor based on the expected net load factor and the relative volatility standard deviation include: using the product of the expected net load factor and the relative volatility standard deviation as a control factor; obtaining the safety margin of the current line based on the expected net load factor, wherein the safety margin is negatively correlated with the expected net load factor; using the ratio of the control factor to the safety margin as a risk index; and using the natural index value of the risk index as the volatility penalty factor.

[0009] The product of the expected net load factor and the standard deviation of relative volatility is used as a control factor, and divided by the safety margin of the line to construct a risk index. This index is then amplified using a natural exponential function to more sensitively reflect risk. Risk only increases sharply when high load factor and high volatility occur simultaneously. Finally, the exponential function amplifies the penalty effect, enabling the algorithm to more accurately identify and avoid extreme risk scenarios, further improving the reliability of decision-making.

[0010] Optionally, the difference between 1 and the expected net load rate can be used as the safety margin for the line.

[0011] Using the difference between 1 and the expected net load rate as the safety margin is simple to calculate and can directly reflect the remaining carrying capacity of the line.

[0012] Optionally, the steps for calculating the relative fluctuation standard deviation include: using the ratio of the standard deviation of historical output data to the average output as the relative fluctuation standard deviation.

[0013] Normalization was performed by dividing it by the average output, thus eliminating the influence of the size of the distributed power generation capacity on the volatility assessment.

[0014] Optionally, the improved ant colony algorithm also includes an improved pheromone update mechanism, which includes: defining the sum of the products of the physical impedance of all lines on the same path and the fluctuation penalty factor as the comprehensive risk cost of the path; and updating the pheromone after each iteration, wherein the increment of the pheromone is inversely proportional to the comprehensive risk cost.

[0015] The overall risk cost is used as the basis for pheromone updates. This cost includes both static physical impedance and dynamic fluctuation risk, so that the algorithm's memory and learning process no longer simply favors the shortest path, but strongly favors the path with lower overall risk.

[0016] Optionally, before obtaining the fluctuation penalty factor, a data acquisition step is also included: acquiring the distribution network topology, line physical impedance, and rated capacity; acquiring the basic load and load to be transferred to the line through the distribution network automation system; and acquiring the historical output data of the distributed power source within a set time period through the distributed power management system or photovoltaic inverter terminal.

[0017] Optional, the duration can be set to 3 to 10 minutes.

[0018] Choosing a short time window of 3 to 10 minutes ensures that the assessed power fluctuations reflect the current, immediate operating state of the power grid, rather than its long-term average. This allows for rapid decision-making in response to drastic changes in power output caused by sudden weather events, improving the timeliness of decision-making.

[0019] Optionally, the formula for calculating the pheromone increment is: In the formula, Ants On the line The increase in pheromones left on the surface; The total amount of pheromones is a constant. For including lines The comprehensive risk cost of the ant's access to the complete path.

[0020] Optionally, it also includes parsing the optimal power transfer path into a set of switch operation instructions and sending them to the distribution network automation system for execution.

[0021] Secondly, this application provides an optimization decision-making system for distribution network transfer paths, which adopts the following technical solution: An optimization decision-making system for distribution network transfer paths includes a processor and a memory, wherein the memory stores computer program instructions, and when the computer program instructions are executed by the processor, the optimization decision-making method for distribution network transfer paths described above is implemented.

[0022] The above-mentioned method for optimizing the distribution network transfer path is generated into a computer program and stored in a memory so that it can be loaded and executed by a processor. Thus, a system can be built based on the memory and processor for easy use.

[0023] This application achieves the following technical effects: It constructs a fluctuation penalty factor that integrates the expected load rate and historical power supply volatility, and then creates dynamic heuristic information based on this factor. This enables the ant colony algorithm to dynamically quantify and avoid potential overload risks caused by unstable renewable energy output. Combined with a pheromone update mechanism based on comprehensive risk cost, it ensures the algorithm can quickly converge to the optimal power transfer path that balances low loss and high reliability, thus improving the robustness and security of the decision-making process. Attached Figure Description

[0024] Figure 1 This is a flowchart of an optimization decision-making method for distribution network transfer paths according to an embodiment of this application.

[0025] Figure 2 This is a flowchart of the method for obtaining the fluctuation penalty factor in an optimization decision-making method for distribution network transfer paths according to an embodiment of this application.

[0026] Figure 3 This is a power distribution network topology diagram in the embodiments of this application.

[0027] Figure 4 This is a power output diagram of the distributed power sources involved in different lines in the embodiments of this application.

[0028] Figure 5 This is a comprehensive risk cost diagram of different paths in the embodiments of this application after iteration by the ant colony algorithm.

[0029] Figure 6 This is a comparison diagram of the load in the line after the power transfer is completed in the embodiment of this application and the line after the power transfer in the traditional transfer path. Detailed Implementation

[0030] This application discloses an optimization decision-making method for power transfer paths in a distribution network. The method calculates the expected net load factor of each line using real-time measurement data, historical output sequences of distributed power sources, and line rated capacity. A fluctuation penalty factor is constructed based on the expected net load factor and the relative fluctuation standard deviation. The fluctuation penalty factor is then used to correct traditional heuristic information to form a dynamic heuristic factor. This enables ants to avoid lines with high load factors and high fluctuations when selecting paths, thereby reducing the risk of line overload after power transfer operations.

[0031] Reference Figure 1 An optimization decision-making method for distribution network transfer paths includes steps S1-S2.

[0032] S1: Based on the improved ant colony algorithm, find the optimal transfer path for the loads to be transferred in the distribution network to connect to a healthy power source.

[0033] In this step, data collection begins.

[0034] The data acquisition steps include: collecting the distribution network topology, line physical impedance, and rated capacity. The current switch status of the entire network, line topology, and the inherent physical parameters of each line are obtained through a Supervisory Control and Data Acquisition (SCADA) system. For example, these parameters include, but are not limited to, the line's physical impedance and rated capacity.

[0035] The basic load and load to be transferred to the line are obtained through the distribution network automation system; Similarly, real-time operational data of the power grid is obtained through SCADA systems or Advanced Measurement Systems (AMIs), including the base load of current customers on each line after removing the output of distributed power sources, to reflect the actual electricity consumption of customers on each line.

[0036] The historical output data of the distributed power source within a set time period is collected through a distributed power management system or a photovoltaic inverter terminal.

[0037] In this embodiment, historical output data of each photovoltaic power station in the distributed power system over the past T minutes are collected through a distributed power management system (DEMS) or a photovoltaic inverter telemetry terminal. The value range can be set to 3 to 10 minutes based on experience; in this embodiment, it is set to minutes.

[0038] After data collection is completed, the dynamic heuristic factor used to guide ants in the improved ant colony algorithm can be calculated; for any line, the dynamic heuristic factor is inversely proportional to the line's physical impedance and fluctuation penalty factor.

[0039] Among them, reference Figure 2 The steps for obtaining the volatility penalty factor include steps S11-S14.

[0040] S11: Calculate the average output of the historical output data of the distributed power sources involved in the line, and take the sum of the line's base load and the load to be transferred in the area to be transferred as the total load; take the ratio of the difference between the total load and the average output of the distributed power sources to the line's rated capacity as the expected net load rate.

[0041] The formula for calculating the expected net load factor can be expressed as: In the formula, Indicates the line The expected net load factor, which is a dimensionless percentage; Indicates the line The basic load on; Indicates the line The average output of the distributed power sources involved is the average of the historical output data collected in step S1. This indicates the amount of load to be transferred within the designated area. Indicates the line The rated capacity is determined by the characteristics of the line itself and is an inherent parameter of the line.

[0042] This represents the line's expected net load factor, which is the net value after the customer's actual electricity demand is partially offset by the average output of photovoltaic power. The closer the expected net load factor is to 1, the stronger the load from the area to be transferred to the line. When the load approaches the rated capacity of the line, it is about to reach the maximum value that the line can withstand, which increases the risk of overload when the load is transferred to the line.

[0043] For example, suppose the rated capacity of a certain line is... At the same time, the basic load of the line was collected. Average output of distributed power sources .

[0044] Ultimately, the expected net load factor The expected net load factor assesses the anticipated stability of the line after the transfer is completed.

[0045] S12: Obtain the relative standard deviation of fluctuation, which represents the degree of fluctuation in historical output data.

[0046] The relative fluctuation standard deviation is mainly used to measure the fluctuation of historical output data of distributed power sources. In this embodiment, the ratio of the standard deviation of historical output data to the average output is used as the relative fluctuation standard deviation.

[0047] The formula for calculating the relative standard deviation of volatility can be expressed as: In the formula, Indicates the line The relative standard deviation of the distributed power sources involved reflects the volatility of historical output data; express Timetable The output of distributed power sources involved; Indicates the line The average output of the distributed power sources involved.

[0048] A higher value indicates a more unstable output from the distributed power source (e.g., in cloudy weather, the value can reach above 0.7); a lower value indicates a more stable output from the distributed power source (e.g., in sunny weather, the value can be below 0.1). This is used in the denominator. It is a robust numerical processing method used to avoid issues with average output. The algorithm's universality is ensured by the fact that division by zero occurs at extremely low times (such as in the early morning or at night).

[0049] S13: Construct a volatility penalty factor based on the expected net load factor and the relative volatility standard deviation. The volatility penalty factor is positively correlated with the expected net load factor and the relative volatility standard deviation.

[0050] Specifically, the product of the expected net load factor and the relative volatility standard deviation is used as a control factor. The safety margin of the current line is obtained based on the expected net load factor, which is negatively correlated with the expected net load factor. The ratio of the control factor to the safety margin is used as a risk index; the natural exponential value of the risk index is used as a volatility penalty factor. The difference between 1 and the expected net load factor is used as the line's safety margin.

[0051] The formula for calculating the volatility penalty factor can be expressed as: In the formula, Indicates the line The volatility penalty factor, which is greater than or equal to 1.0; Indicates the line Expected net load factor; Indicates the line The relative standard deviation of the fluctuation; Represents the natural exponential function; This is a hyperparameter, mainly used to prevent the denominator from being 0, and can be set to 0.1.

[0052] This section combines the expected net load factor and the relative standard deviation of volatility. This item will only increase significantly when both a high expected load factor and a high relative standard deviation of volatility occur simultaneously. This represents the remaining available capacity of the line, and the safety margin approaches 0 when the expected net load factor approaches 1.0.

[0053] Instance-like, , Safety margin is The final calculated volatility penalty factor is: .

[0054] Similarly, suppose another route , .

[0055] Its volatility penalty factor As can be seen, although the load factor differs by only 0.2, the risk penalty factor differs by about 40 times (from 2.26 to 90.03). This accurately reflects the threat of volatility under high load, and the use of an exponential function effectively amplifies this risk.

[0056] Finally, a volatility penalty factor is used to correct the heuristic information in the relevant technology to obtain a dynamic heuristic factor.

[0057] The dynamic heuristic factor for any line is inversely proportional to both the line's physical impedance and the fluctuation penalty factor.

[0058] The formula for calculating the dynamic heuristic factor can be expressed as: In the formula, Indicates the line A dynamic heuristic factor is used to guide the ants' path selection; Indicates the physical impedance of the line; Indicates the line The volatility penalty factor.

[0059] A line A high dynamic heuristic factor requires both low physical impedance and a low volatility penalty factor to be met simultaneously. If a path has low physical impedance but a high volatility penalty factor, its calculated overall cost will be low. The value remains high, resulting in a smaller dynamic heuristic factor, and ants will automatically avoid this high-risk path.

[0060] The improved ant colony algorithm also includes an improved pheromone update mechanism, which includes: defining the sum of the products of the physical impedance of all lines on the same path and the fluctuation penalty factor as the comprehensive risk cost of the path; and updating the pheromone after each iteration, wherein the increment of the pheromone is inversely proportional to the comprehensive risk cost.

[0061] The improved ant colony algorithm in this application rewards paths with low overall risk and cost.

[0062] For any complete path traversed by any ant, the sum of the total costs of all segments along that complete path is taken as the total risk cost of that path.

[0063] In the formula, Indicates the first The complete path traveled by an ant The overall risk cost; Indicates the first The complete path traversed by a single ant; Indicates the physical impedance of the line; Indicates the line The volatility penalty factor.

[0064] Indicates the line The overall cost is used to measure the quality of the route. The lower the overall risk cost of the complete route, the lower the cost and risk of the route.

[0065] The pheromone on the path is updated based on the comprehensive risk cost of the entire path, for any given route. The pheromone increment on the surface can be calculated using the following formula: ; Ants In the path The increase in pheromones left on the surface; Represents the total pheromone constant (e.g.) ); Indicates the first The total risk cost of the complete path traversed by an ant.

[0066] Once the dynamic heuristic factor of the circuit and the pheromone update mechanism are determined, the algorithm can be initialized and then started.

[0067] Initialization settings include setting the number of ants. (For example Maximum number of iterations (For example ), pheromone importance factor (For example ), the importance of dynamic heuristic factors (For example ), pheromone evaporation rate (For example ) and pheromone constant .

[0068] Then repeat the execution. This is the second iteration. Each iteration includes the following steps: Path building: Combining Figure 3 , Each ant starts from a node awaiting load transfer and, based on dynamic heuristic factors and pheromone update mechanisms, independently chooses a path to a healthy power source (the power source). or power supply The path of ).

[0069] Cost calculation: Calculate the overall risk cost of the path chosen by each ant.

[0070] Pheromone Update: After all ants have completed their search, the global pheromone level is updated according to the pheromone update mechanism.

[0071] Decision output: After each iteration, the algorithm converges. The path with the highest global pheromone concentration or the lowest overall risk cost during the iteration process is selected as the final optimal pheromone transfer path.

[0072] Combination Figure 4 and Figure 5 It can be seen that the path The output fluctuations of the distributed power sources involved are relatively small, therefore the overall risk cost calculated for this path is higher, and the final path is determined accordingly. This is the optimal transfer path. Compared to the traditional ant colony algorithm, which only considers the impedance in the path, since the impedance of path A is less than that of path B, the traditional ant colony algorithm ultimately selects path A. Combined with... Figure 6 Because the distributed power supply processing involved in path A has large fluctuations, the subsequent power supply load has large fluctuations, which poses a risk of overload leading to secondary failures.

[0073] S2: Parse the optimal power transfer path into a set of switch operation instructions and send them to the distribution network automation system for execution.

[0074] The optimal power transfer path is parsed into a set of switch operation instructions, such as: closing the tie switch K5, opening the fault switch K2, and sending it to the SCADA system for execution to complete the power transfer operation after a power supply failure in the distribution network.

[0075] This application also discloses an optimization decision-making system for distribution network transfer paths, including a processor and a memory. The memory stores computer program instructions, and when the computer program instructions are executed by the processor, an optimization decision-making method for distribution network transfer paths according to this application is implemented.

[0076] The system also includes other components well known to those skilled in the art, such as communication buses and communication interfaces, the settings and functions of which are known in the art and will not be described in detail here.

[0077] The above are all preferred embodiments of this application, and are not intended to limit the scope of protection of this application. Therefore, all equivalent changes made in accordance with the structure, shape and principle of this application should be covered within the scope of protection of this application.

Claims

1. An optimization decision-making method for distribution network transfer paths, characterized in that, include: An improved ant colony algorithm is used to find the optimal transfer path for loads to be transferred in the distribution network to connect to a healthy power source. The improved ant colony algorithm includes a dynamic heuristic factor to guide the ants; for any path, the dynamic heuristic factor is inversely proportional to the path's physical impedance and fluctuation penalty factor. The steps for obtaining the fluctuation penalty factor include: calculating the average output of the historical output data of the distributed power sources involved in the line; taking the sum of the line's base load and the load to be transferred in the area to be transferred as the total load; taking the ratio of the difference between the total load and the average output of the distributed power sources to the line's rated capacity as the expected net load rate; obtaining the relative fluctuation standard deviation that characterizes the degree of fluctuation in historical output data; constructing the fluctuation penalty factor based on the expected net load rate and the relative fluctuation standard deviation; and the fluctuation penalty factor being positively correlated with the expected net load rate and the relative fluctuation standard deviation.

2. The method for optimizing distribution network transfer paths according to claim 1, characterized in that, The steps for constructing a volatility penalty factor based on the expected net load factor and the relative volatility standard deviation include: using the product of the expected net load factor and the relative volatility standard deviation as a control factor; obtaining the safety margin of the current line based on the expected net load factor, which is negatively correlated with the expected net load factor; using the ratio of the control factor to the safety margin as a risk index; and using the natural index value of the risk index as the volatility penalty factor.

3. The method for optimizing distribution network transfer paths according to claim 2, characterized in that, The difference between 1 and the expected net load rate is used as the safety margin of the line.

4. The method for optimizing distribution network transfer paths according to claim 1, characterized in that, The steps for calculating the relative fluctuation standard deviation include: using the ratio of the standard deviation of historical output data to the average output as the relative fluctuation standard deviation.

5. The method for optimizing distribution network transfer paths according to claim 1, characterized in that, The improved ant colony algorithm also includes an improved pheromone update mechanism, which includes: defining the sum of the products of the physical impedance of all lines on the same path and the fluctuation penalty factor as the comprehensive risk cost of the path; and updating the pheromone after each iteration, wherein the increment of the pheromone is inversely proportional to the comprehensive risk cost.

6. The method for optimizing distribution network transfer paths according to claim 1, characterized in that, Before the step of obtaining the fluctuation penalty factor, there is also a data acquisition step: acquiring the topology, physical impedance and rated capacity of the distribution network; acquiring the basic load and load to be transferred to the line through the distribution network automation system; and acquiring the historical output data of the distributed power source within a set time period through the distributed power management system or photovoltaic inverter terminal.

7. The method for optimizing distribution network transfer paths according to claim 6, characterized in that, Set the duration to 3-10 minutes.

8. The method for optimizing distribution network transfer paths according to claim 5, characterized in that, The formula for calculating the pheromone increment is: In the formula, Ants On the line The increase in pheromones left on the surface; The total amount of pheromones is a constant. For including lines The comprehensive risk cost of the ant's access to the complete path.

9. The method for optimizing distribution network transfer paths according to claim 1, characterized in that, It also includes parsing the optimal power transfer path into a set of switch operation instructions and sending them to the distribution network automation system for execution.

10. An optimization decision-making system for power distribution network transfer paths, characterized in that, include: A processor and a memory, wherein the memory stores computer program instructions that, when executed by the processor, implement an optimization decision-making method for distribution network transfer paths according to any one of claims 1-9.

Images