Optimized dispatching method and system for power distribution network containing distributed power supply

Abstract

The invention relates to the technical field of power distribution scheduling, in particular to an optimal scheduling method and system for a power distribution network containing a distributed power supply, and the method comprises the steps: S1, collecting the data of the distributed power network, and S2, carrying out the real-time analysis of the data of the distributed power network, according to the method, a target transfer decision is obtained through the step S3, an execution result is obtained through the step S4, the state of a node surge risk assessment value is judged through the step S5, the abrupt change duration state of single data in distributed power grid data is judged through the step S6, and the abrupt change duration state of the single data in the distributed power grid data is judged through the step S5. According to the method, the influence of the fragile nodes on data optimization is reasonably reduced through the step S7, the influence of the fragile nodes and the surge risk assessment value on data optimization in the target transfer decision is reasonably reduced through the step S8, and therefore the efficiency of power distribution network dispatching optimization and the feeder automation accuracy are improved.

Description

Technical Field

[0001] This invention relates to the field of power distribution dispatching technology, and in particular to a method and system for optimizing the dispatching of power distribution networks containing distributed power sources. Background Technology

[0002] In the field of power distribution dispatching technology, the complexity of operation and control has increased significantly with the large-scale integration of distributed generation sources. Traditional and current mainstream methods for optimizing and dispatching distribution networks with distributed generation sources generally rely on historical data and ideal models, lacking high-precision, real-time closed-loop feedback on the actual effects of dispatching commands. This insufficient accuracy and delay in the feedback loop prevents the system from accurately perceiving the real-time impact of dispatching nodes, making critical lines or equipment susceptible to damage. Furthermore, the failure to perform rapid closed-loop correction based on the actual grid conditions reduces the iterative efficiency of the optimization dispatching process, thus limiting the overall optimization effect of distribution networks with distributed generation sources.

[0003] Chinese patent CN113472014B discloses a method for optimizing the scheduling of distribution networks containing distributed generation. This method includes: collecting data, market information, and uncertainty information from various nodes of distributed generation in the distribution network across different regions; preprocessing this data and storing it in a data information cloud platform; calling the data information cloud platform and, based on the system's multi-objective attributes, constructing an optimized scheduling model of the distribution network based on the economic benefits, losses, and generation costs of distributed generation, according to constraints; using the constraints as decision analysis conditions, outputting the predicted data for each region and the specific range of uncertainty fluctuations; inputting the corresponding operating schemes and uncertainties into multiple optimization objective functions to calculate the overall benefit; and having decision-makers in different regions compare the overall benefit of the schemes and their own regional development status to select the best strategy as the optimized scheduling scheme. However, this method still suffers from insufficient accuracy in feeder automation, low efficiency in balancing scheduling, and vulnerability of weak nodes in distribution networks containing distributed generation, resulting in low efficiency in optimized scheduling. Summary of the Invention

[0004] To address this, the present invention provides a method and system for optimizing the scheduling of distribution networks containing distributed power sources, thereby overcoming the problems of insufficient accuracy of feeder automation and low efficiency of balance scheduling in the prior art, which leads to vulnerable nodes in distribution networks containing distributed power sources being easily damaged and low efficiency of distribution network optimization scheduling.

[0005] To achieve the above objectives, the present invention provides a method for optimized scheduling of a distribution network containing distributed generation sources, the method comprising: Step S1: Collect distributed power grid data; Step S2: Perform real-time analysis on the distributed power grid data to obtain the target analysis data; Step S3: Obtain target mobilization decisions based on target analysis data; Step S4: Execute the mobilization decision based on the target and obtain the execution result; Step S5: Obtain the real-time surge risk assessment result value based on the execution result, obtain the vulnerable node based on the real-time surge risk assessment result value, and optimize the decision based on the acquisition process of the target mobilization decision based on the vulnerable node. Step S6: Perform predictive analysis based on distributed power grid data to obtain a set of data mutation durations, and optimize the target analysis data based on the set of data mutation durations; Step S7: Based on the data mutation duration set, determine the vulnerable nodes, obtain the vulnerable node determination results, and update the data in the data optimization process based on the vulnerable node determination results. Step S8 involves correcting the data during the data optimization process based on the target mobilization decision, and also supplementing the data during the data optimization process based on the real-time surge risk assessment results.

[0006] Further, in step S2, when performing real-time analysis on the distributed power grid data according to the real-time analysis method, the real-time analysis method includes: Step A01: The grid storage capacity, grid real-time power and grid real-time current are used as grid real-time data; the node real-time power, node real-time current and node storage capacity are used as node real-time data; the grid charging and discharging status, grid peak electricity consumption and grid peak power generation are used as grid status data; and the node charging and discharging status, node peak electricity consumption and node peak power generation are used as node status data. Step A02: Perform state encoding processing on the power grid state data and node state data respectively to obtain encoded power grid state data and encoded node state data; Step A03: Output the real-time power grid data, real-time node data, encoded power grid status data, and encoded node status data as target analysis data.

[0007] Further, in step S3, when the target mobilization decision acquisition method acquires the target mobilization decision based on the target analysis data, the target mobilization decision acquisition method includes: Step B01: Set the scheduling objectives; Step B02: Input the scheduling achievement purpose and target analysis data into the scheduling decision generation model to obtain the scheduling decision set output by the scheduling decision generation model; Step B03: Sort the scheduling decision set to obtain a recommended sorting series; Step B04: The first scheduling decision in the recommended ranking series is taken as the target dispatch decision. The target dispatch decision includes the scheduling object, time point, scheduling quota, and scheduling role.

[0008] Furthermore, in step S4, the mobilization is executed according to the target mobilization decision, and the execution result is obtained.

[0009] Further, in step S5, when obtaining the real-time surge risk assessment result value based on the execution result, the execution result is input into the risk assessment model to obtain the surge risk real-time assessment result value output by the risk assessment model. The surge risk real-time assessment result value includes the node surge risk assessment value my and the system stability assessment value hy. In step S5, when acquiring vulnerable nodes based on the real-time surge risk assessment result, the node surge risk assessment value my is compared with the preset node surge risk assessment value my0. The state of the node surge risk assessment value is determined based on the comparison result, and the vulnerable node is acquired based on the determination result, wherein: When my≤my0, the surge risk assessment value of the node is determined to be normal, and the vulnerable node is not acquired; When my > my0, the node surge risk assessment value is determined to be abnormal, the node is output as a vulnerable node, and the process of obtaining the target mobilization decision is optimized. In step S5, when optimizing the target mobilization decision acquisition process, the system stability evaluation value hy is compared with the preset system stability evaluation value hy0. Based on the comparison result, the state of the system stability evaluation value is judged, and based on the judgment result, the target mobilization decision acquisition process is optimized, wherein: When hy≥hy0, the system stability evaluation value is determined to be normal, and the process of obtaining the target mobilization decision is optimized: the second scheduling decision in the recommended ranking series is taken as the target mobilization decision, and the mobilization is re-executed based on the target mobilization decision; When hy < hy0, the system stability evaluation value is determined to be abnormal, and the process of obtaining the target mobilization decision is optimized: the scheduling to achieve the goal is manually reset to obtain the reset scheduling to achieve the goal, and the target mobilization decision is obtained again based on the reset scheduling to achieve the goal.

[0010] Further, in step S6, when performing predictive analysis based on distributed grid data, the distributed grid data is input into the mutation duration prediction model to obtain the set of data mutation durations tp={tp1,tp2,...,tpf} output by the mutation duration prediction model, where tp1 is the first data mutation duration in the set of data mutation durations, tp2 is the second data mutation duration in the set of data mutation durations, tpf is the f-th data mutation duration in the set of data mutation durations, f is the order of the data mutation durations, and f is a positive integer, the maximum value of which is the total number of distributed grid data; In step S6, when optimizing the target analysis data based on the data mutation duration set, the first data mutation duration tp1 is compared with the preset data mutation duration tp0. The state of the first data mutation duration is determined based on the comparison result, and the target analysis data is optimized based on the determination result, wherein: When tp1≥tp0, the duration of the first data mutation is considered acceptable, and no data optimization is performed on the target analysis data. When tp1 < tp0, the duration of the first data mutation is deemed unacceptable, and data optimization is performed on the target analysis data: The single data stability ys is obtained, compared with a preset single data stability ys0, and the state of single data stability is judged based on the comparison result. Based on the judgment result, data optimization is performed on the target analysis data, wherein: When ys≥ys0, the stability of a single data point is determined to be stable, and data optimization is performed on the target analysis data: the first data point is marked as a stability mutation data point, and the stability mutation data point is added to the target analysis data; When ys < ys0, the stability of the single data is determined to be unstable. Data optimization is performed on the target analysis data: the first data is manually replaced to obtain the first data after replacement, and the first data after replacement is added to the target analysis data. The second data mutation duration tp2 is compared with the preset data mutation duration tp0. Based on the comparison result, the state of the second data mutation duration is determined, and the target analysis data is optimized based on the determination result. Specifically: When tp2≥tp0, the duration of the second data mutation is considered acceptable, and no data optimization is performed on the target analysis data. When tp2 < tp0, the duration of the second data mutation is deemed unacceptable, and data optimization is performed on the target analysis data: The single data stability ys is obtained, compared with a preset single data stability ys0, and the state of single data stability is judged based on the comparison result. Based on the judgment result, data optimization is performed on the target analysis data, wherein: When ys≥ys0, the stability of a single data point is determined to be stable. Data optimization is then performed on the target analysis data: the second data point is marked as a stability mutation data point, and the stability mutation data point is added to the target analysis data. When ys < ys0, the stability of the single data is determined to be unstable. Data optimization is performed on the target analysis data: the second data is manually replaced to obtain the first data after replacement, and the first data after replacement is added to the target analysis data. ... The duration of the f-th data mutation, tpf, is compared with the preset duration of the data mutation, tp0. Based on the comparison result, the state of the duration of the f-th data mutation is determined, and the target analysis data is optimized based on the determination result. Wherein: When tpf≥tp0, the duration of the f-th data mutation is considered acceptable, and no data optimization is performed on the target analysis data. When tpf < tp0, the duration of the f-th data mutation is deemed unacceptable, and data optimization is performed on the target analysis data: The single data stability ys is obtained, compared with a preset single data stability ys0, and the state of single data stability is judged based on the comparison result. Based on the judgment result, data optimization is performed on the target analysis data, wherein: When ys≥ys0, the stability of a single data point is determined to be stable, and data optimization is performed on the target analysis data: the f-th data point is marked as a stability mutation data point, and the stability mutation data point is added to the target analysis data. When ys < ys0, the stability of the single data is determined to be unstable. Data optimization is performed on the target analysis data: the f-th data is manually replaced to obtain the first data after replacement, and the first data after replacement is added to the target analysis data.

[0011] Furthermore, in step S7, the vulnerable nodes are determined from the data mutation duration set, and the data optimization process is updated based on the vulnerable node determination result. When a single node in the distributed power grid data is not a vulnerable node, the data optimization process is not updated. When a single node in the distributed power grid data is a vulnerable node, the data optimization process involves data updating: the preset data mutation duration tp0 is updated according to the data update coefficient cv, and the updated preset data mutation duration is set to tp01, where tp01 = tp0 × cv. The updated preset data mutation duration tp01 is output as the preset data mutation duration tp0, and the first data mutation duration tp1, the second data mutation duration tp2, and the fth data mutation duration tpf are re-compared with the preset data mutation duration tp0.

[0012] Furthermore, in step S8, data correction is performed on the data optimization process based on the target mobilization decision. When the scheduling object in the target mobilization decision does not contain a vulnerable node, no data correction is performed on the data optimization process. When the scheduling object in the target mobilization decision contains vulnerable nodes, data correction is performed during the data optimization process: Specifically, in step S8, data correction is performed on the data optimization process based on the real-time surge risk assessment result value. The preset data mutation duration tp0 is corrected according to the data correction coefficient cn. The corrected preset data mutation duration is set as tp02, where tp02 = tp0 × cn. The updated preset data mutation duration tp02 is output as the preset data mutation duration tp0. The first data mutation duration tp1, the second data mutation duration tp2, and the fth data mutation duration tpf are then compared with the preset data mutation duration tp0 again.

[0013] Further, in step S8, data correction is performed on the data optimization process based on the real-time surge risk assessment result. The node surge risk assessment value my is compared with the preset node surge risk assessment value my0. The state of the node surge risk assessment value is judged based on the comparison result, and data correction is performed on the data optimization process based on the judgment result. Wherein: When my≤my0, the state of the node surge risk assessment value is determined to be normal, and no data correction is performed during the data optimization process; When my > my0, the node surge risk assessment value is determined to be abnormal. Data correction is then performed during the data optimization process based on the data correction coefficient ca, set as ca = 1.38 - 0.22 × e. -（my-my0） Where e is the base of the natural logarithm, the preset data mutation duration after correction is set to tp03, tp03 = tp0 × ca, the updated preset data mutation duration tp03 is output as the preset data mutation duration tp0, and the first data mutation duration tp1, the second data mutation duration tp2 and the fth data mutation duration tpf are re-compared with the preset data mutation duration tp0.

[0014] On the other hand, the present invention also provides a system for optimizing the scheduling of a distribution network containing distributed power sources, the system comprising: The data acquisition module is used to collect data from the distributed power grid. The real-time analysis module is used to perform real-time analysis on distributed power grid data to obtain target analysis data; The scheduling decision module is used to obtain target deployment decisions based on target analysis data; The scheduling and execution module is used to execute the dispatch based on the target dispatch decision and obtain the execution result; The execution feedback module is used to obtain the real-time surge risk assessment result value based on the execution result, to obtain the vulnerable node based on the real-time surge risk assessment result value, and to optimize the decision based on the acquisition process of the target mobilization decision based on the vulnerable node. The predictive analysis module is used to perform predictive analysis based on distributed power grid data, obtain a set of data mutation durations, optimize the target analysis data based on the set of data mutation durations, identify vulnerable nodes based on the set of data mutation durations, obtain the vulnerability node identification results, update the data optimization process based on the vulnerability node identification results, correct the data optimization process based on the target mobilization decision, and supplement the data optimization process based on the real-time surge risk assessment results.

[0015] Compared with the prior art, the beneficial effects of the present invention are as follows: the method collects distributed power grid data in step S1 to facilitate subsequent optimized scheduling based on the distributed power grid data; the method also performs real-time analysis of the distributed power grid data in step S2, converting the textual status data in the distributed power grid data into numerical representation to improve model recognition efficiency; the method further obtains the target dispatch decision in step S3 to quickly determine the optimal scheduling decision to achieve the objective at the current moment; the method further obtains the execution result in step S4 to facilitate subsequent optimization of vulnerable nodes based on the execution result; and the method further judges the status of the node surge risk assessment value in step S5 to assess vulnerable nodes. The method acquires and judges the status of system stability assessment values ​​to optimize the target dispatch decision acquisition process in real time based on the system stability, thereby improving the efficiency of distribution network balance dispatch. The method also judges the mutation duration status of individual data in the distributed grid data and the stability status of individual data in step S6, so as to reasonably optimize the specific mutation data in the distributed grid data. The method also reasonably reduces the impact of vulnerable nodes on data optimization in step S7. The method also reasonably reduces the impact of vulnerable nodes and surge risk assessment values ​​on data optimization in target dispatch decision in step S8, thereby protecting vulnerable nodes and improving the efficiency of distribution network dispatch optimization and the accuracy of feeder automation. Attached Figure Description

[0016] Figure 1 This is a flowchart illustrating the distribution network optimization scheduling method with distributed power sources in this embodiment; Figure 2 This is a flowchart illustrating the real-time analysis method in this embodiment; Figure 3 This is a flowchart illustrating the target mobilization decision acquisition method in this embodiment; Figure 4 This is a schematic diagram of the system structure of the distribution network optimization scheduling method with distributed power sources in this embodiment. Detailed Implementation

[0017] To make the objectives and advantages of the present invention clearer, the present invention will be further described below with reference to embodiments; it should be understood that the specific embodiments described herein are merely for explaining the present invention and are not intended to limit the present invention.

[0018] Preferred embodiments of the present invention will now be described with reference to the accompanying drawings. Those skilled in the art should understand that these embodiments are merely illustrative of the technical principles of the present invention and are not intended to limit the scope of protection of the present invention.

[0019] It should be noted that in the description of this invention, the terms "upper," "lower," "left," "right," "inner," and "outer," etc., which indicate directions or positional relationships, are based on the directions or positional relationships shown in the accompanying drawings. This is merely for ease of description and does not indicate or imply that the device or element must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, it should not be construed as a limitation of this invention.

[0020] Furthermore, it should be noted that, in the description of this invention, unless otherwise explicitly specified and limited, the terms "installation," "connection," and "joining" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral connection; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; and they can refer to the internal connection of two components. Those skilled in the art can understand the specific meaning of the above terms in this invention according to the specific circumstances.

[0021] Please see Figure 1 As shown, this is a flowchart illustrating the distribution network optimization scheduling method with distributed generation in this embodiment. The method includes: Step S1: Collect distributed power grid data; Step S2: Perform real-time analysis on the distributed power grid data to obtain the target analysis data; Step S3: Obtain target mobilization decisions based on target analysis data; Step S4: Execute the mobilization decision based on the target and obtain the execution result; Step S5: Obtain the real-time surge risk assessment result value based on the execution result, obtain the vulnerable node based on the real-time surge risk assessment result value, and optimize the decision based on the acquisition process of the target mobilization decision based on the vulnerable node. Step S6: Perform predictive analysis based on distributed power grid data to obtain a set of data mutation durations, and optimize the target analysis data based on the set of data mutation durations; Step S7: Based on the data mutation duration set, determine the vulnerable nodes, obtain the vulnerable node determination results, and update the data in the data optimization process based on the vulnerable node determination results. Step S8 involves correcting the data during the data optimization process based on the target mobilization decision, and also supplementing the data during the data optimization process based on the real-time surge risk assessment results.

[0022] Specifically, the method for optimizing the scheduling of distribution networks containing distributed power sources is applied to a distributed power source distribution network scheduling terminal, such as the power distribution equipment in resettlement housing. The method collects distributed power network data, obtains target dispatch decisions based on this data, and optimizes the data based on the execution results to protect vulnerable nodes. Specifically, the method collects distributed power network data in step S1 to facilitate subsequent optimized scheduling. Step S2 involves real-time analysis of the distributed power network data, converting textual status data into numerical representations to improve model recognition efficiency. Step S3 involves obtaining target dispatch decisions to quickly determine the optimal scheduling decision for achieving the desired outcome at the current moment. Step S4 involves obtaining the execution results to facilitate subsequent optimization. The method optimizes vulnerable nodes based on the execution results. Step S5 further determines the state of the node surge risk assessment value to identify vulnerable nodes and determines the state of the system stability assessment value. This allows for real-time optimization of the target dispatch decision acquisition process based on system stability, thereby improving the efficiency of distribution network balancing and dispatching. Step S6 further determines the duration of sudden changes in individual data within the distributed grid data and assesses the stability of individual data, enabling reasonable optimization of specific sudden changes in the distributed grid data. Step S7 further reduces the impact of vulnerable nodes on data optimization. Step S8 further reduces the impact of vulnerable nodes and surge risk assessment values ​​on data optimization in the target dispatch decision, thereby protecting vulnerable nodes and improving the efficiency of distribution network dispatch optimization and the accuracy of feeder automation.

[0023] Specifically, in step S1, distributed grid data is collected. The distributed grid data includes grid energy storage, grid real-time power, grid real-time current, grid charging and discharging status, node real-time power, node real-time current, node energy storage, node charging and discharging status, and peak status. The peak status includes grid electricity consumption peak, grid power generation peak, node electricity consumption peak, and node power generation peak.

[0024] Specifically, the grid storage capacity refers to the total electrical energy currently stored in the distribution network; the grid real-time power refers to the electrical power of the distribution network at the current moment; the grid real-time current refers to the current quantity and direction of current transmission in the distribution network at the current moment; the grid charging / discharging state refers to the power exchange relationship between the distribution network as a whole and the external main grid; the node real-time power refers to the electrical power of a single node at the current moment; the node real-time current refers to the current quantity and direction of current transmission in a single node at the current moment; the node storage capacity refers to the total electrical energy currently stored in a single node in the distribution network; and the node charging / discharging state refers to the current inflow / outflow state of a single node in the distribution network, with current flowing into a single node considered as charging and current flowing out of a single node considered as discharging. A single node refers to edge power supply equipment in the distribution network, such as energy storage batteries and small wind turbines. This embodiment does not refer to the grid. The specific methods for obtaining the stored power, real-time grid power, grid charging and discharging status, node real-time power, node stored power, and node charging and discharging status are not limited. Those skilled in the art can freely choose according to actual needs, such as by installing smart meters and power sensors at individual nodes and user main inlets. The peak power consumption of the grid refers to the period with the highest electricity demand in the distribution network, the peak power generation of the grid refers to the period with the highest power generation in the distribution network, the peak power consumption of a node refers to the period with the highest electricity demand of a single node, and the peak power generation of a node refers to the period with the highest power generation of a single node. This embodiment does not limit the specific methods for obtaining the peak power consumption of the grid, the peak power generation of the grid, the peak power consumption of a node, and the peak power generation of a node. Those skilled in the art can freely choose according to actual needs, such as obtaining the peak status through the energy management system of the distribution network.

[0025] Specifically, in step S1, distributed power grid data is collected to facilitate subsequent optimized scheduling based on the distributed power grid data, thereby improving power distribution efficiency.

[0026] Specifically, in step S2, when performing real-time analysis on the distributed power grid data according to the real-time analysis method, the real-time analysis method includes: Step A01: The grid storage capacity, grid real-time power and grid real-time current are used as grid real-time data; the node real-time power, node real-time current and node storage capacity are used as node real-time data; the grid charging and discharging status, grid peak electricity consumption and grid peak power generation are used as grid status data; and the node charging and discharging status, node peak electricity consumption and node peak power generation are used as node status data. Step A02: Perform state encoding processing on the power grid state data and node state data respectively to obtain encoded power grid state data and encoded node state data; Step A03: Output the real-time power grid data, real-time node data, encoded power grid status data, and encoded node status data as target analysis data.

[0027] Specifically, the state encoding process refers to the process of representing the textual state data in the power grid state data and node state data using digital encoding. This embodiment does not limit the specific method of state encoding processing. Those skilled in the art can freely choose according to actual needs. For example, the power grid charging / discharging state is defined as the number 1 when it is charging, the power grid charging / discharging state is defined as the number 2 when it is discharging, the peak power consumption of the power grid is defined as a, the peak power generation of the power grid is defined as b, the node charging / discharging state is defined as 3 when it is charging, the node charging / discharging state is defined as 4 when it is discharging, the peak power consumption of the node is defined as c, and the peak power generation of the node is defined as d.

[0028] Specifically, in step S2, the distributed power grid data is analyzed in real time, and the textual status data in the distributed power grid data is converted into digital representation to improve the model recognition efficiency, thereby improving the optimization and scheduling efficiency of the distribution network.

[0029] Specifically, in step S3, when the target mobilization decision acquisition method acquires a target mobilization decision based on target analysis data, the target mobilization decision acquisition method includes: Step B01: Set the scheduling objectives; Step B02: Input the scheduling achievement purpose and target analysis data into the scheduling decision generation model to obtain the scheduling decision set output by the scheduling decision generation model; Step B03: Sort the scheduling decision set to obtain a recommended sorting series; Step B04: The first scheduling decision in the recommended ranking series is taken as the target dispatch decision. The target dispatch decision includes the scheduling object, time point, scheduling quota, and scheduling role.

[0030] Specifically, the "dispatch achievement objective" refers to the target state to be achieved after dispatching the distribution network, such as maximizing the economic benefits of the distribution network. The "dispatch decision generation model" is a deep learning model that takes the dispatch achievement objective and target analysis data as input data and the dispatch decision set as output data. This embodiment does not limit the specific construction method of the dispatch decision generation model; those skilled in the art can freely choose according to actual needs. For example, historical dispatch achievement objective and target analysis data can be used as historical dispatch data, and the historical dispatch data and its corresponding dispatch decision set can be used as a training set to train the deep learning model to obtain the dispatch decision generation model. The dispatch decision set refers to the set of data generated based on the dispatch decision generation model. The set of multiple scheduling decisions obtained from the model refers to the specific scheme for scheduling the distribution network to achieve the scheduling objective. The recommended ranking refers to the sequential list of all scheduling decisions in the set of scheduling decisions arranged from best to worst according to the scheduling objective. For example, all scheduling decisions in the set of scheduling decisions can be sorted from largest to smallest according to economic benefits. The scheduling object refers to the single node that needs to be scheduled in the target scheduling decision. The time point refers to the specific moment when the single node is scheduled in the target scheduling decision. The scheduling quota refers to the specific amount of electricity to be scheduled for the single node at the time point in the target scheduling decision. The scheduling role refers to the method of scheduling the single node, such as charging and discharging.

[0031] Specifically, in step S3, the target dispatch decision is obtained in order to quickly determine the best dispatch decision to achieve the goal at the current moment, thereby improving the efficiency of distribution network dispatch optimization.

[0032] Specifically, in step S4, the mobilization is carried out according to the target mobilization decision, and the execution result is obtained.

[0033] Specifically, the dispatch execution refers to the process of the distribution network executing the target dispatch decision, and the execution result refers to the distributed power grid data re-collected after executing the target dispatch decision.

[0034] Specifically, in step S4, the execution result is obtained so that vulnerable nodes can be optimized based on the execution result, thereby improving the optimization scheduling efficiency.

[0035] Specifically, in step S5, when obtaining the real-time surge risk assessment result value based on the execution result, the execution result is input into the risk assessment model to obtain the real-time surge risk assessment result value output by the risk assessment model. The real-time surge risk assessment result value includes the node surge risk assessment value my and the system stability assessment value hy. In step S5, when acquiring vulnerable nodes based on the real-time surge risk assessment result, the node surge risk assessment value my is compared with the preset node surge risk assessment value my0. The state of the node surge risk assessment value is determined based on the comparison result, and the vulnerable node is acquired based on the determination result, wherein: When my≤my0, the surge risk assessment value of the node is determined to be normal, and the vulnerable node is not acquired; When my > my0, the node surge risk assessment value is determined to be abnormal, the node is output as a vulnerable node, and the process of obtaining the target mobilization decision is optimized. In step S5, when optimizing the target mobilization decision acquisition process, the system stability evaluation value hy is compared with the preset system stability evaluation value hy0. Based on the comparison result, the state of the system stability evaluation value is judged, and based on the judgment result, the target mobilization decision acquisition process is optimized, wherein: When hy≥hy0, the system stability evaluation value is determined to be normal, and the process of obtaining the target mobilization decision is optimized: the second scheduling decision in the recommended ranking series is taken as the target mobilization decision, and the mobilization is re-executed based on the target mobilization decision; When hy < hy0, the system stability evaluation value is determined to be abnormal, and the process of obtaining the target mobilization decision is optimized: the scheduling to achieve the goal is manually reset to obtain the reset scheduling to achieve the goal, and the target mobilization decision is obtained again based on the reset scheduling to achieve the goal.

[0036] Specifically, the risk assessment model refers to a recurrent neural network model that takes the execution results as input data and the real-time surge risk assessment result value as output data. This embodiment does not limit the specific construction method of the risk assessment model; those skilled in the art can freely choose according to actual needs. For example, the recurrent neural network model can be trained using historical execution results and their corresponding real-time surge risk assessment result values ​​as a training set to obtain the risk assessment model. The node surge risk assessment value refers to the surge risk assessment value of a single node obtained according to the risk assessment model. The surge risk assessment value is a numerical value that measures the probability of the node suffering surge damage. The system stability assessment value is a numerical value obtained according to the risk assessment model that measures the overall stability of the distribution network system at the current moment. The preset node surge risk assessment value is a preset value used to judge the state of the node surge risk assessment value. This embodiment does not limit the specific value setting of the preset node surge risk assessment value my0; those skilled in the art can freely choose according to actual needs. For example, in this embodiment, m is set according to the required accuracy of distribution network risk monitoring. y0=0.23, the state of the node surge risk assessment value refers to the degree of normality of the node surge risk assessment value judged based on the node surge risk assessment value and the preset node surge risk assessment value. The state of the node surge risk assessment value includes normal and abnormal. The preset system stability assessment value refers to the preset value used to judge the state of the system stability assessment value. This embodiment does not limit the specific value setting of the preset system stability assessment value hy0. Those skilled in the art can freely choose according to actual needs. For example, this embodiment sets hy0=0.85 based on historical experience. The state of the system stability assessment value refers to the degree of normality of the system stability assessment value judged based on the system stability assessment value and the preset system stability assessment value. The state of the system stability assessment value includes normal and abnormal. The manual reset refers to the process by which relevant technicians reset the dispatching to achieve the purpose according to the current state of the distribution network. This embodiment does not limit the specific method of manual reset. Those skilled in the art can freely choose according to actual needs. For example, relevant technicians can reset the dispatching to achieve the purpose through the system interaction interface.

[0037] Specifically, in step S5, the status of the node surge risk assessment value is judged to identify vulnerable nodes, and the status of the system stability assessment value is judged to optimize the target dispatch decision acquisition process in real time based on the system stability, thereby improving the efficiency of distribution network balance dispatch.

[0038] Specifically, in step S6, when performing predictive analysis based on distributed grid data, the distributed grid data is input into the mutation duration prediction model to obtain the data mutation duration set tp={tp1,tp2,...,tpf} output by the mutation duration prediction model, where tp1 is the first data mutation duration in the data mutation duration set, tp2 is the second data mutation duration in the data mutation duration set, tpf is the f-th data mutation duration in the data mutation duration set, f is the order of the data mutation durations, and f is a positive integer, the maximum value of which is the total number of distributed grid data; In step S6, when optimizing the target analysis data based on the data mutation duration set, the first data mutation duration tp1 is compared with the preset data mutation duration tp0. The state of the first data mutation duration is determined based on the comparison result, and the target analysis data is optimized based on the determination result, wherein: When tp1≥tp0, the duration of the first data mutation is considered acceptable, and no data optimization is performed on the target analysis data. When tp1 < tp0, the duration of the first data mutation is deemed unacceptable, and data optimization is performed on the target analysis data: The single data stability ys is obtained, compared with a preset single data stability ys0, and the state of single data stability is judged based on the comparison result. Based on the judgment result, data optimization is performed on the target analysis data, wherein: When ys≥ys0, the stability of a single data point is determined to be stable, and data optimization is performed on the target analysis data: the first data point is marked as a stability mutation data point, and the stability mutation data point is added to the target analysis data; When ys < ys0, the stability of the single data is determined to be unstable. Data optimization is performed on the target analysis data: the first data is manually replaced to obtain the first data after replacement, and the first data after replacement is added to the target analysis data. The second data mutation duration tp2 is compared with the preset data mutation duration tp0. Based on the comparison result, the state of the second data mutation duration is determined, and the target analysis data is optimized based on the determination result. Specifically: When tp2≥tp0, the duration of the second data mutation is considered acceptable, and no data optimization is performed on the target analysis data. When tp2 < tp0, the duration of the second data mutation is deemed unacceptable, and data optimization is performed on the target analysis data: The single data stability ys is obtained, compared with a preset single data stability ys0, and the state of single data stability is judged based on the comparison result. Based on the judgment result, data optimization is performed on the target analysis data, wherein: When ys≥ys0, the stability of a single data point is determined to be stable. Data optimization is then performed on the target analysis data: the second data point is marked as a stability mutation data point, and the stability mutation data point is added to the target analysis data. When ys < ys0, the stability of the single data is determined to be unstable. Data optimization is performed on the target analysis data: the second data is manually replaced to obtain the first data after replacement, and the first data after replacement is added to the target analysis data. ... The duration of the f-th data mutation, tpf, is compared with the preset duration of the data mutation, tp0. Based on the comparison result, the state of the duration of the f-th data mutation is determined, and the target analysis data is optimized based on the determination result. Wherein: When tpf≥tp0, the duration of the f-th data mutation is considered acceptable, and no data optimization is performed on the target analysis data. When tpf < tp0, the duration of the f-th data mutation is deemed unacceptable, and data optimization is performed on the target analysis data: The single data stability ys is obtained, compared with a preset single data stability ys0, and the state of single data stability is judged based on the comparison result. Based on the judgment result, data optimization is performed on the target analysis data, wherein: When ys≥ys0, the stability of a single data point is determined to be stable, and data optimization is performed on the target analysis data: the f-th data point is marked as a stability mutation data point, and the stability mutation data point is added to the target analysis data. When ys < ys0, the stability of the single data is determined to be unstable. Data optimization is performed on the target analysis data: the f-th data is manually replaced to obtain the first data after replacement, and the first data after replacement is added to the target analysis data.

[0039] Specifically, the mutation duration prediction model refers to a deep learning model that takes distributed grid data as input data and a set of data mutation durations as output data. This embodiment does not limit the specific construction method of the mutation duration prediction model; those skilled in the art can freely choose according to actual needs. For example, historical distributed grid data and its corresponding set of data mutation durations can be used as a training set to train the deep learning model to obtain the mutation duration prediction model. The set of data mutation durations refers to the time length of mutations in all individual data points in the distributed grid data obtained according to the mutation duration prediction model. The first data mutation duration refers to the time length of mutation of the first data point in the distributed grid data. The duration of the second data mutation refers to the time length during which the second data in the distributed grid data undergoes a mutation. The duration of the f-th data mutation refers to the time length during which the f-th data in the distributed grid data undergoes a mutation. The total number of distributed grid data refers to the total number of all data in the distributed grid data. The preset data mutation duration refers to a preset value for judging the state of a single data mutation duration. This embodiment does not limit the specific value of the preset data mutation duration tp0. Those skilled in the art can freely choose according to actual needs. For example, this embodiment sets tp0 = 10 minutes based on historical experience. The state of the first data mutation duration refers to the judgment based on the first data mutation duration and the preset data mutation duration. The acceptable duration of the first data mutation is defined as either acceptable or unacceptable. Single data stability refers to a numerical value measuring the stability of the data. This embodiment does not limit the specific method for obtaining single data stability; those skilled in the art can freely choose according to actual needs. For example, single data stability can be obtained through expert evaluation, which refers to the process by which experts with the ability to set single data stability parameters set the stability. This embodiment does not limit the specific method by which experts set single data stability; for example, experts can input single data stability through a cloud-based interactive window. Preset single data stability refers to judging the state of single data stability. The preset value for single data stability ys0 is not limited in this embodiment. Those skilled in the art can freely choose the value according to actual needs, such as setting ys0=80 in a percentage system. The state of single data stability refers to the degree of stability of a single data point judged based on its stability and the preset single data stability. The state of single data stability includes stable and unstable. Manual replacement refers to the process by which relevant technicians manually replace the data. The state of the second data mutation duration refers to the acceptable degree of the second data mutation duration judged based on its stability and the preset data mutation duration. The state of the second data mutation duration includes acceptable and unacceptable.The state of the duration of the f-th data mutation refers to the acceptability of the duration of the f-th data mutation, judged based on its comparison with a preset duration. The state of the duration of the f-th data mutation includes acceptable and unacceptable.

[0040] Specifically, in step S6, the duration of mutations in individual data in the distributed power grid data is judged, and the stability of individual data is judged, so as to reasonably optimize the specific mutation data in the distributed power grid data, thereby improving the efficiency and accuracy of scheduling optimization.

[0041] Specifically, in step S7, the vulnerable nodes are determined from the data mutation duration set, and the data optimization process is updated based on the vulnerable node determination results. When a single node in the distributed power grid data is not a vulnerable node, the data optimization process is not updated. When a single node in the distributed power grid data is a vulnerable node, the data optimization process involves data updating: the preset data mutation duration tp0 is updated according to the data update coefficient cv, and the updated preset data mutation duration is set to tp01, where tp01 = tp0 × cv. The updated preset data mutation duration tp01 is output as the preset data mutation duration tp0, and the first data mutation duration tp1, the second data mutation duration tp2, and the fth data mutation duration tpf are re-compared with the preset data mutation duration tp0.

[0042] Specifically, this embodiment does not limit the specific value of the data update coefficient cv. Those skilled in the art can freely choose according to actual needs, as long as the need to reasonably increase the preset data mutation time tp0 is met. For example, this embodiment sets cv=1.33 based on historical experience.

[0043] Specifically, in step S7, when a single node in the distributed power grid data is a vulnerable node, the preset data mutation duration is increased according to the data update coefficient in order to reasonably reduce the impact of vulnerable nodes on data optimization, thereby improving the efficiency of distribution network scheduling optimization and the accuracy of feeder automation.

[0044] Specifically, in step S8, data correction is performed on the data optimization process based on the target mobilization decision. When the scheduling object in the target mobilization decision does not contain a vulnerable node, no data correction is performed on the data optimization process. When the scheduling object in the target mobilization decision contains vulnerable nodes, data correction is performed during the data optimization process: Specifically, in step S8, data correction is performed on the data optimization process based on the real-time surge risk assessment result value. The preset data mutation duration tp0 is corrected according to the data correction coefficient cn. The corrected preset data mutation duration is set as tp02, where tp02 = tp0 × cn. The updated preset data mutation duration tp02 is output as the preset data mutation duration tp0. The first data mutation duration tp1, the second data mutation duration tp2, and the fth data mutation duration tpf are then compared with the preset data mutation duration tp0 again.

[0045] Specifically, this embodiment does not limit the specific value of the data correction coefficient cn. Those skilled in the art can freely choose according to actual needs, as long as the need to reasonably increase the preset data mutation duration tp0 is met. For example, this embodiment sets cv=1.36 based on historical experience.

[0046] Specifically, in step S8, when the scheduling object in the target dispatch decision contains a vulnerable node, the preset data mutation duration is increased according to the data correction coefficient, so as to reasonably reduce the impact of the vulnerable node in the target dispatch decision on data optimization, thereby improving the efficiency of distribution network dispatch optimization and the accuracy of feeder automation.

[0047] Specifically, in step S8, the data optimization process is corrected based on the real-time surge risk assessment result. The node surge risk assessment value my is compared with the preset node surge risk assessment value my0. The state of the node surge risk assessment value is judged based on the comparison result, and the data optimization process is corrected based on the judgment result. Wherein: When my≤my0, the state of the node surge risk assessment value is determined to be normal, and no data correction is performed during the data optimization process; When my > my0, the node surge risk assessment value is determined to be abnormal. Data correction is then performed during the data optimization process based on the data correction coefficient ca, set as ca = 1.38 - 0.22 × e. -（my-my0） Where e is the base of the natural logarithm, the preset data mutation duration after correction is set to tp03, tp03 = tp0 × ca, the updated preset data mutation duration tp03 is output as the preset data mutation duration tp0, and the first data mutation duration tp1, the second data mutation duration tp2 and the fth data mutation duration tpf are re-compared with the preset data mutation duration tp0.

[0048] Specifically, in step S8, when the state of the node surge risk assessment value is abnormal, the preset data mutation duration is increased according to the data correction coefficient which increases from 1.16 to 1.38. Here, the constant term 1.38 of the data correction coefficient is the maximum achievable value of the data correction coefficient, and the constant coefficient 1.22 is the change range of the data correction coefficient. This is to reasonably reduce the impact of the surge risk assessment value on data optimization, thereby protecting vulnerable nodes and improving the efficiency of distribution network dispatch optimization and the accuracy of feeder automation.

[0049] Please see Figure 4 As shown, this is a schematic diagram of the system structure of the distribution network optimization scheduling method with distributed generation in this embodiment. The system includes: The data acquisition module is used to collect data from the distributed power grid. A real-time analysis module is used to perform real-time analysis on distributed power grid data to obtain target analysis data. The real-time analysis module is connected to the data acquisition module. The scheduling decision module is used to obtain target deployment decisions based on target analysis data, and the scheduling decision module is connected to the real-time analysis module; The scheduling execution module is used to execute the scheduling decision based on the target scheduling decision and obtain the execution result. The scheduling execution module is connected to the scheduling decision module. The execution feedback module is used to obtain the real-time surge risk assessment result value based on the execution result, and to obtain the vulnerable node based on the real-time surge risk assessment result value. It is also used to optimize the decision based on the acquisition process of the target mobilization decision of the vulnerable node. The execution feedback module is connected to the scheduling execution module. The predictive analysis module is used to perform predictive analysis based on distributed power grid data to obtain a set of data mutation durations, optimize the target analysis data based on the set of data mutation durations, identify vulnerable nodes based on the set of data mutation durations, obtain the identification results, update the data optimization process based on the identification results, correct the data optimization process based on the target mobilization decision, and supplement the data optimization process based on the real-time surge risk assessment results. The predictive analysis module is connected to the data acquisition module.

[0050] The technical solution of the present invention has been described above with reference to the preferred embodiments shown in the accompanying drawings. However, it will be readily understood by those skilled in the art that the scope of protection of the present invention is obviously not limited to these specific embodiments. Without departing from the principles of the present invention, those skilled in the art can make equivalent changes or substitutions to the relevant technical features, and the technical solutions after these changes or substitutions will all fall within the scope of protection of the present invention.

Claims

1. A method for optimal dispatch of a power distribution network with distributed power sources, characterized in that, The method comprises: Step S1, collecting distributed power grid data; Step S2, performing real-time analysis on the distributed power grid data to obtain target analysis data; Step S3, obtaining a target mobilization decision based on the target analysis data; Step S4, performing mobilization execution based on the target mobilization decision to obtain an execution result; Step S5, obtaining a surge risk real-time evaluation result value based on the execution result, and obtaining a vulnerable node based on the surge risk real-time evaluation result value, and optimizing the target mobilization decision based on the vulnerable node; Step S6, performing prediction analysis based on the distributed power grid data to obtain a data mutation duration set, and optimizing the target analysis data based on the data mutation duration set; Step S7, judging the vulnerable node based on the data mutation duration set to obtain a vulnerable node judgment result, and updating the data optimization process based on the vulnerable node judgment result; Step S8, correcting the data optimization process based on the target mobilization decision, and correcting the data optimization process based on the surge risk real-time evaluation result value.

2. The optimal scheduling method of the distribution network with distributed power sources according to claim 1, characterized in that, In the step S2, when the real-time analysis method is used to analyze the distributed power grid data in real time, the real-time analysis method comprises: Step A01, taking the power grid storage capacity, the power grid real-time power and the power grid real-time current as power grid real-time data, taking the node real-time power, the node real-time current and the node storage capacity as node real-time data, taking the power grid charging and discharging state, the power grid power consumption peak and the power grid power generation peak as power grid state data, and taking the node charging and discharging state, the node power consumption peak and the node power generation peak as node state data; Step A02, performing state coding processing on the power grid state data and the node state data respectively to obtain coded power grid state data and coded node state data; Step A03, outputting the power grid real-time data, the node real-time data, the coded power grid state data and the coded node state data as target analysis data.

3. The method of claim 2, wherein, In the step S3, when the target mobilization decision is obtained based on the target analysis data by the target mobilization decision obtaining method, the target mobilization decision obtaining method comprises: Step B01, setting a dispatching purpose; Step B02, inputting the dispatching purpose and the target analysis data into a dispatching decision generation model to obtain a dispatching decision set output by the dispatching decision generation model; Step B03, recommending and sorting the dispatching decision set to obtain a recommended sorting series; Step B04, taking the first dispatching decision in the recommended sorting series as the target mobilization decision, the target mobilization decision comprising a dispatching object, a time point, a dispatching quota and a dispatching role.

4. The method of claim 3, wherein, In the step S4, the execution result is obtained based on the target mobilization decision.

5. The method of claim 4, wherein, In the step S5, when the execution result is input into a risk evaluation model to obtain a surge risk real-time evaluation result value output by the risk evaluation model, the surge risk real-time evaluation result value comprises a node surge risk evaluation value my and a system stability evaluation value hy. In the step S5, when the vulnerable node is acquired according to the real-time surge risk evaluation result value, the node surge risk evaluation value my and the preset node surge risk evaluation value my0 are compared, the state of the node surge risk evaluation value is judged according to the comparison result, and the vulnerable node is acquired according to the judgment result, wherein: When my≤my0, it is determined that the state of the node surge risk evaluation value is normal, and the vulnerable node is not acquired; When my>my0, it is determined that the state of the node surge risk evaluation value is abnormal, the node is output as a vulnerable node, and the decision optimization of the acquisition process of the target mobilization decision is performed; In the step S5, when the decision optimization of the acquisition process of the target mobilization decision is performed, the system stability evaluation value hy and the preset system stability evaluation value hy0 are compared, the state of the system stability evaluation value is judged according to the comparison result, and the decision optimization of the acquisition process of the target mobilization decision is performed according to the judgment result, wherein: When hy≥hy0, it is determined that the state of the system stability evaluation value is normal, the decision optimization of the acquisition process of the target mobilization decision is performed: the second scheduling decision in the recommended sorting series is taken as the target mobilization decision, and the mobilization execution is re-performed according to the target mobilization decision; When hy 6. The method of claim 5, wherein, In the step S6, when the prediction analysis is performed according to the distributed power grid data, the distributed power grid data is input into the mutation time length prediction model, and a data mutation time length set tp={tp1, tp2,..., tpf} output by the mutation time length prediction model is obtained, wherein tp1 is the first data mutation time length in the data mutation time length set, tp2 is the second data mutation time length in the data mutation time length set, tpf is the fth data mutation time length in the data mutation time length set, f is the order of the data mutation time length, and f is a positive integer, and the maximum value is the total number of the distributed power grid data; In the step S6, when the target analysis data is data optimized according to the data mutation time length set, the first data mutation time length tp1 and the preset data mutation time length tp0 are compared, the state of the first data mutation time length is judged according to the comparison result, and the target analysis data is data optimized according to the judgment result, wherein: When tp1≥tp0, it is determined that the state of the first data mutation time length is acceptable, and the target analysis data is not data optimized; When tp1 When tp1 When ys≥ ys0, it is determined that the state of single data stability is stable, and the target analysis data is optimized: the first data is marked as stability mutation data, and the stability mutation data is supplemented into the target analysis data; When ys< ys0, it is determined that the state of single data stability is unstable, and the target analysis data is optimized: the first data is manually replaced to obtain the first data after replacement, and the first data after replacement is added to the target analysis data; The second data mutation time length tp2 is compared with the preset data mutation time length tp0, the state of the second data mutation time length is judged according to the comparison result, and the target analysis data is optimized according to the judgment result, wherein: When tp2≥ tp0, it is determined that the state of the second data mutation time length is acceptable, and the target analysis data is not optimized; When tp2< tp0, it is determined that the state of the second data mutation time length is unacceptable, and the target analysis data is optimized: the single data stability ys is obtained, the single data stability ys is compared with the preset single data stability ys0, the state of the single data stability is judged according to the comparison result, and the target analysis data is optimized according to the judgment result, wherein: When ys≥ ys0, it is determined that the state of single data stability is stable, and the target analysis data is optimized: the second data is marked as stability mutation data, and the stability mutation data is supplemented into the target analysis data; When ys< ys0, it is determined that the state of single data stability is unstable, and the target analysis data is optimized: the second data is manually replaced to obtain the first data after replacement, and the first data after replacement is added to the target analysis data; …… The fth data mutation time length tpf is compared with the preset data mutation time length tp0, the state of the fth data mutation time length is judged according to the comparison result, and the target analysis data is optimized according to the judgment result, wherein: When tpf≥ tp0, it is determined that the state of the fth data mutation time length is acceptable, and the target analysis data is not optimized; When tpf< tp0, it is determined that the state of the fth data mutation time length is unacceptable, and the target analysis data is optimized: the single data stability ys is obtained, the single data stability ys is compared with the preset single data stability ys0, the state of the single data stability is judged according to the comparison result, and the target analysis data is optimized according to the judgment result, wherein: When ys≥ ys0, it is determined that the state of single data stability is stable, and the target analysis data is optimized: the fth data is marked as stability mutation data, and the stability mutation data is supplemented into the target analysis data; When ys< ys0, it is determined that the state of single data stability is unstable, and the target analysis data is optimized: the fth data is manually replaced to obtain the first data after replacement, and the first data after replacement is added to the target analysis data.

7. The method of claim 6, wherein, In the step S7, the fragile node is judged according to the data mutation time set, and the data updating is performed on the data optimization process according to the fragile node judgment result; when the single node in the distributed power grid data is not the fragile node, the data updating is not performed on the data optimization process; When the single node in the distributed power grid data is the fragile node, the data updating is performed on the data optimization process: the preset data mutation time tp0 is updated according to the data updating coefficient cv, the updated preset data mutation time is set as tp01, tp01=tp0×cv, the updated preset data mutation time tp01 is output as the preset data mutation time tp0, and the first data mutation time tp1, the second data mutation time tp2 and the fth data mutation time tpf are compared with the preset data mutation time tp0 again.

8. The method of claim 7, wherein, In the step S8, the data correction is performed on the data optimization process according to the target mobilization decision; when the scheduling object in the target mobilization decision does not contain the fragile node, the data correction is not performed on the data optimization process; When the scheduling object in the target mobilization decision contains the fragile node, the data correction is performed on the data optimization process: Specifically, in the step S8, the data correction is performed on the data optimization process according to the surge risk real-time evaluation result value; the preset data mutation time tp0 is corrected according to the data correction coefficient cn, the corrected preset data mutation time is set as tp02, tp02=tp0×cn, the updated preset data mutation time tp02 is output as the preset data mutation time tp0, and the first data mutation time tp1, the second data mutation time tp2 and the fth data mutation time tpf are compared with the preset data mutation time tp0 again.

9. The method of claim 8, wherein, In the step S8, the data correction is performed on the data optimization process according to the surge risk real-time evaluation result value; the node surge risk evaluation value my and the preset node surge risk evaluation value my0 are compared, the state of the node surge risk evaluation value is judged according to the comparison result, and the data correction is performed on the data optimization process according to the judgment result, wherein: When my≤my0, it is determined that the state of the node surge risk evaluation value is normal, and the data correction is not performed on the data optimization process; When my>my0, it is determined that the state of the surge risk assessment value of the node is abnormal, a data correction process is performed on the data optimization according to a data correction coefficient ca, and ca=1.38-0.22×e is set -（my-my0） wherein e is the base of natural logarithm, tp03=tp0×ca is set as the preset data mutation duration after correction, the updated preset data mutation duration tp03 is output as the preset data mutation duration tp0, and the first data mutation duration tp1, the second data mutation duration tp2, and the fth data mutation duration tpf are compared with the preset data mutation duration tp0 again.

10. A system applied to the distributed power grid optimization scheduling method of any one of claims 1-9, the system comprising: a data acquisition module configured to acquire distributed power grid data; a real-time analysis module configured to perform real-time analysis on the distributed power grid data to obtain target analysis data; a scheduling decision module configured to obtain a target mobilization decision according to the target analysis data; a scheduling execution module configured to perform mobilization execution according to the target mobilization decision to obtain an execution result; an execution feedback module configured to obtain a surge risk real-time evaluation result value according to the execution result, and obtain a fragile node according to the surge risk real-time evaluation result value, and further configured to optimize the process of obtaining the target mobilization decision according to the fragile node. The prediction analysis module is configured to perform prediction analysis on the distributed power grid data to obtain a data mutation duration set, perform data optimization on target analysis data according to the data mutation duration set, perform a vulnerable node judgment according to the data mutation duration set to obtain a vulnerable node judgment result, perform data updating on the data optimization process according to the vulnerable node judgment result, perform data correction on the data optimization process according to a target mobilization decision, and perform data correction on the data optimization process according to a surge risk real-time evaluation result value.

Images