CN117455269B

CN117455269B - Snowflake type power distribution network power supply safety prediction method, device, equipment and storage medium

Info

Publication number: CN117455269B
Application number: CN202311770637.4A
Authority: CN
Inventors: 王哲; 段佳莉; 张章; 罗涛; 罗凤章; 王伟臣; 姚宗强; 迟福建; 王海波; 吴璇; 朱林伟; 葛楠; 李娟�
Original assignee: State Grid Corp of China SGCC; State Grid Tianjin Electric Power Co Ltd; Chengnan Power Supply Co of State Grid Tianjin Electric Power Co Ltd
Current assignee: State Grid Corp of China SGCC; State Grid Tianjin Electric Power Co Ltd; Chengnan Power Supply Co of State Grid Tianjin Electric Power Co Ltd
Priority date: 2023-12-21
Filing date: 2023-12-21
Publication date: 2024-04-16
Anticipated expiration: 2043-12-21
Also published as: CN117455269A

Abstract

The embodiment of the invention provides a snowflake type power distribution network power supply safety prediction method, device, equipment and storage medium, wherein the method comprises the following steps: determining at least one target index of the snowflake type power distribution network and at least one target feature corresponding to the target index, wherein the target index comprises an index influencing the power supply safety of the snowflake type power distribution network, and the target feature comprises a power factor influencing the target index; generating a causal relationship graph according to the target index, the target feature and the relationship between the target index and the target feature; establishing a dynamics model for predicting power supply safety according to the causal relationship graph; according to the dynamic model, the power supply safety condition of the snowflake type power distribution network in a future preset time period is predicted, the technical problem that the prediction accuracy of the power supply safety condition of the snowflake type power distribution network in the prior art is low is solved, and the technical effect of accurately predicting the power supply safety condition of the snowflake type power distribution network is achieved.

Description

Snowflake type power distribution network power supply safety prediction method, device, equipment and storage medium

Technical Field

The invention belongs to the technical field of electric power, and particularly relates to a snowflake type power distribution network power supply safety prediction method, device, equipment and storage medium.

Background

The snowflake type power distribution network power supply safety management is a key for ensuring the operation safety of a power grid and the power consumption safety of a user, the power consumption quality and safety of the user are directly affected when the snowflake type power distribution network is used as the last kilometer of a power supply system, and once accidents occur, equipment damage and power failure can be caused, and the life and property safety of related personnel can be threatened, so that the power supply safety management of the snowflake type power distribution network is particularly important.

In the prior art, single indexes of a snowflake type power distribution network are subjected to trend translation, curve fitting and other modes based on historical data to obtain predicted values of power supply safety conditions of the snowflake type power distribution network in the future year, the prediction accuracy is poor, the future power supply safety conditions are predicted based on random forest method, vector machine and other modes, but the algorithms are more suitable for short-term prediction and static prediction, and the relation among a plurality of influencing factors in a power supply system can be ignored, so that the prediction accuracy is low.

Therefore, the technical problem of low prediction accuracy of the power supply safety condition of the snowflake type power distribution network exists in the prior art.

Disclosure of Invention

The invention provides a snowflake type power distribution network power supply safety prediction method, device, equipment and storage medium, and aims to solve the technical problem that in the prior art, the accuracy of predicting the power supply safety condition of a snowflake type power distribution network is low.

In a first aspect, the invention provides a snowflake type power distribution network power supply safety prediction method, which comprises the following steps:

determining at least one target index of the snowflake type power distribution network and at least one target feature corresponding to the target index, wherein the target index comprises an index influencing the power supply safety of the snowflake type power distribution network, and the target feature comprises a power factor influencing the target index;

generating a causal relationship graph according to the target index, the target feature and the relationship between the target index and the target feature;

establishing a dynamics model for predicting power supply safety according to the causal relationship graph;

And predicting the power supply safety condition of the snowflake type power distribution network in a preset time period in the future according to the dynamics model.

Further, determining at least one target indicator of the snowflake-type power distribution network includes:

Acquiring at least one initial index affecting power supply safety, and generating a fishbone graph according to the initial index, wherein the fishbone graph is used for combing the layers of the initial index;

And determining target indexes according to the fishbone diagram, wherein the target indexes comprise a capacity-to-load ratio and a load transfer rate.

Further, determining at least one target feature corresponding to the target index includes:

generating a first matrix according to the target index;

Generating a second matrix according to at least one initial feature affecting the target index and the initial feature;

generating a PLS regression model according to the first matrix and the second matrix;

Calculating a VIP score of each initial feature according to the PLS regression model;

and determining target characteristics according to the VIP score.

Further, generating a causal relationship graph according to the target index, the target feature, and the relationship between the target index and the target feature, including:

taking the target index and the target characteristic as nodes, wherein the nodes comprise affected nodes and affected nodes;

a causal relationship graph is obtained by connecting nodes with relationships by using edges with arrows, wherein the directions of the arrows point to affected factors from the affected factors.

Further, according to the causal relation graph, a dynamics model for predicting power supply safety is established, which comprises the following steps:

Generating a stack flow graph according to the causal relation graph;

According to a stack flow diagram, a first sub-model for predicting the capacity ratio is established, wherein the first sub-model comprises a first state variable equation, a first speed equation and a first auxiliary equation, and the first state variable equation is as follows: ,/> Is the value of the load transfer rate/> time, is the value of the load transfer rate initial time/> ,/> is the rate of change of the load transfer rate at time t, is the input rate of the load transfer rate,/> is the output rate of the load transfer rate, and the first rate equation is: the auxiliary variable value of the load transfer rate is given by/(,/>), the exogenous variable value of the load transfer rate is given by , the linear function is given by/(), and the first auxiliary equation is: ,/> Is an auxiliary variable other than the auxiliary variable/> of the load transfer rate;

According to the stack flow graph, a second sub-model for predicting the load transfer rate is established, wherein the second sub-model comprises a second state variable equation, a second rate equation and a second auxiliary equation, and the second state variable equation is as follows: , Is the value of the capacity ratio/> , the value of the capacity ratio/> is the value of the initial time/> , the value of the capacity ratio/> is the rate at which the capacity ratio changes at the time t, the value of the capacity ratio/> is the input rate of the capacity ratio, the value of the capacity ratio/> is the output rate of the capacity ratio, and the second rate equation is: the value of the auxiliary variable for the capacitance ratio,/> ,/>, the value of the exogenous variable for the capacitance ratio,/> , and the second auxiliary equation is: ,/> Is an auxiliary variable other than the auxiliary variable/> of the capacitance ratio.

Further, according to the dynamics model, predicting the power supply safety condition of the snowflake type power distribution network in a preset time period in the future includes:

Predicting the capacity-to-load ratio and the load transfer rate of the snowflake type power distribution network in a preset time period in the future;

And predicting the power supply safety condition of the snowflake type power distribution network in a preset time period in the future according to the capacity-to-load ratio and the load transfer rate of the snowflake type power distribution network in the preset time period in the future.

In a second aspect, the present invention provides a snowflake type power distribution network power supply safety prediction device, including:

the determining module is used for determining at least one target index of the snowflake-type power distribution network and at least one target characteristic corresponding to the target index, wherein the target index comprises an index for influencing the power supply safety of the snowflake-type power distribution network, and the target characteristic comprises a power factor for influencing the target index;

the generation module is used for generating a causal relationship graph according to the target index, the target feature and the relationship between the target index and the target feature;

The building module is used for building a dynamics model for predicting power supply safety according to the causal relationship graph;

And the prediction module is used for predicting the power supply safety condition of the snowflake type power distribution network in a preset time period in the future according to the dynamics model.

Further, the determining module comprises an acquiring unit and a first determining unit;

The device comprises an acquisition unit, a control unit and a control unit, wherein the acquisition unit is used for acquiring at least one initial index affecting the power supply safety and generating a fishbone diagram according to the initial index, wherein the fishbone diagram is used for combing the layers of the initial index;

and the first determining unit is used for determining target indexes according to the fishbone diagram, wherein the target indexes comprise a capacity-to-load ratio and a load transfer rate.

Further, the determining module further comprises a first generating unit, a second generating unit, a third generating unit, a calculating unit and a second determining unit;

the first generation unit is used for generating a first matrix according to the target index;

the second generation unit is used for generating a second matrix according to at least one initial characteristic affecting the target index and the initial characteristic;

The third generation unit is used for generating a PLS regression model according to the first matrix and the second matrix;

A calculation unit for calculating VIP score of each initial feature according to PLS regression model;

And the second determining unit is used for determining the target characteristics according to the VIP score.

Further, the generating module comprises a node determining unit and an edge generating unit;

A node determining unit, configured to take a target index and a target feature as nodes, where the nodes include an affected node and an affected node;

And the side generating unit is used for connecting nodes with relations by using the sides with the arrows to obtain a causal relation graph, wherein the directions of the arrows point to the affected factors from the affected factors.

Further, the building module comprises a fourth generating unit, a first building unit and a second building unit;

a fourth generating unit, configured to generate a stack flow graph according to the causal relationship graph;

The first building unit is used for building a first sub-model for predicting the capacity ratio according to the stack flow diagram, wherein the first sub-model comprises a first state variable equation, a first speed equation and a first auxiliary equation, and the first state variable equation is as follows: ,/> Is the value of the load transfer rate/> time, is the value of the load transfer rate initial time/> ,/> is the rate of change of the load transfer rate at time t, is the input rate of the load transfer rate,/> is the output rate of the load transfer rate, and the first rate equation is: the auxiliary variable value of the load transfer rate is given by/(,/>), the exogenous variable value of the load transfer rate is given by , the linear function is given by/(), and the first auxiliary equation is: ,/> Is an auxiliary variable other than the auxiliary variable/> of the load transfer rate;

The second building unit is used for building a second sub-model for predicting the load transfer rate according to the stack flow graph, wherein the second sub-model comprises a second state variable equation, a second rate equation and a second auxiliary equation, and the second state variable equation is: , Is the value of the capacity ratio/> , the value of the capacity ratio/> is the value of the initial time/> , the value of the capacity ratio/> is the rate at which the capacity ratio changes at the time t, the value of the capacity ratio/> is the input rate of the capacity ratio, the value of the capacity ratio/> is the output rate of the capacity ratio, and the second rate equation is: the value of the auxiliary variable for the capacitance ratio,/> ,/>, the value of the exogenous variable for the capacitance ratio,/> , and the second auxiliary equation is: ,/> Is an auxiliary variable other than the auxiliary variable/> of the capacitance ratio.

Further, the prediction module comprises a first prediction unit and a second prediction unit;

the first prediction unit is used for predicting the capacity-to-load ratio and the load transfer rate of the snowflake-type power distribution network in a preset time period in the future;

the second prediction unit is used for predicting the power supply safety condition of the snowflake type power distribution network in a preset time period in the future according to the capacity ratio and the load transfer rate of the snowflake type power distribution network in the preset time period in the future.

In a third aspect, the present invention provides an electronic device, comprising:

one or more processors;

A memory for storing one or more programs;

The one or more programs, when executed by the one or more processors, cause the one or more processors to implement the snowflake-type power distribution network power supply safety prediction method as described in the first aspect.

In a fourth aspect, the present invention also provides a computer readable storage medium having stored thereon a computer program which when executed by a processor implements a snowflake power distribution network power supply security prediction method according to the first aspect.

The embodiment of the invention provides a snowflake type power distribution network power supply safety prediction method, device, equipment and storage medium, wherein the method comprises the following steps: determining at least one target index of the snowflake-type power distribution network and at least one target characteristic corresponding to the target index, wherein the target index comprises an index influencing the power supply safety of the snowflake-type power distribution network, the target characteristic comprises a power factor influencing the target index, and the future power supply safety of the snowflake-type power distribution network is predicted according to the index influencing the power supply safety of the snowflake-type power distribution network and the main power factor corresponding to the index, so that the prediction accuracy is improved; generating a causal relationship graph according to the target index, the target feature and the relationship between the target index and the target feature, and fully considering the interrelationship between each index and the factors by carrying out hierarchical division on the target index and the influencing factors, thereby improving the prediction accuracy; according to the causal relationship graph, a dynamic model for predicting power supply safety is established, the superiority of the dynamic model in dynamic data prediction and long-term prediction is utilized, and the prediction accuracy is improved; according to the dynamic model, the power supply safety condition of the snowflake type power distribution network in a future preset time period is predicted, the technical problem that the prediction accuracy of the power supply safety condition of the snowflake type power distribution network in the prior art is low is solved, and the technical effect of accurately predicting the power supply safety condition of the snowflake type power distribution network is achieved.

Additional features and advantages of the invention will be set forth in the description which follows, and in part will be obvious from the description, or may be learned by practice of the invention. The objectives and other advantages of the invention may be realized and attained by the structure particularly pointed out in the written description and claims hereof as well as the appended drawings.

Drawings

In order to more clearly illustrate the embodiments of the present invention or the technical solutions of the prior art, the following description will briefly explain the drawings used in the embodiments or the description of the prior art, and it is obvious that the drawings in the following description are some embodiments of the present invention, and other drawings can be obtained according to these drawings without inventive effort for a person skilled in the art.

Fig. 1 is a schematic flow chart of a snowflake type power distribution network power supply safety prediction method provided by an embodiment of the invention;

FIG. 2 is a schematic diagram of a level of initial indicators combed by a fishbone graph according to an embodiment of the present invention;

FIG. 3 is a schematic diagram of determining target characteristics by PLS-VIP method according to an embodiment of the present invention;

FIG. 4 is a schematic diagram of a relationship between load transfer rates and corresponding target features according to an embodiment of the present invention;

FIG. 5 is a schematic diagram of the relationship between the capacity ratio and the corresponding target feature according to the embodiment of the present invention;

FIG. 6 is a schematic diagram of a causal relationship graph provided by an embodiment of the present invention;

FIG. 7 is a schematic diagram of a stack flow diagram provided by an embodiment of the present invention;

FIG. 8 is a graph showing the comparison of predicted and actual values of load transfer rates in the verification areas 2010-2020 according to an embodiment of the present invention;

Fig. 9 is a schematic structural diagram of a snowflake type power distribution network power supply safety prediction device according to an embodiment of the present invention;

fig. 10 is a schematic structural diagram of an electronic device according to an embodiment of the present invention.

Detailed Description

For the purpose of making the objects, technical solutions and advantages of the embodiments of the present invention more apparent, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention, and it is apparent that the described embodiments are some embodiments of the present invention, but not all embodiments of the present invention. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.

The following disclosure provides many different embodiments, or examples, for implementing different structures of the invention. In order to simplify the present disclosure, components and arrangements of specific examples are described below. They are, of course, merely examples and are not intended to limit the invention. Furthermore, the present invention may repeat reference numerals and/or letters in the various examples. This repetition is for the purpose of simplicity and clarity and does not in itself dictate a relationship between the various embodiments and/or configurations discussed.

In order to solve the technical problem of low prediction accuracy of power supply safety conditions of a snowflake type power distribution network in the prior art, the embodiment of the invention provides a power supply safety prediction method of a snowflake type power distribution network, as shown in fig. 1: fig. 1 is a schematic flow chart of power supply safety prediction of a snowflake type power distribution network, which is provided by the embodiment of the invention, and the method comprises the following steps:

S101, determining at least one target index of a snowflake type power distribution network and at least one target feature corresponding to the target index, wherein the target index comprises an index influencing the power supply safety of the snowflake type power distribution network, and the target feature comprises a power factor influencing the target index;

s102, generating a causal relationship graph according to target indexes, target characteristics and the relationship between the target indexes and the target characteristics;

S103, establishing a dynamics model for predicting power supply safety according to the causal relationship graph;

S104, predicting the power supply safety condition of the snowflake type power distribution network in a preset time period in the future according to the dynamics model.

According to the embodiment of the invention, at least one target index of the snowflake-type power distribution network and at least one target characteristic corresponding to the target index are determined, wherein the target index comprises an index influencing the power supply safety of the snowflake-type power distribution network, the target characteristic comprises a power factor influencing the target index, and the future power supply safety of the snowflake-type power distribution network is predicted according to the index influencing the power supply safety of the snowflake-type power distribution network and the main power factor corresponding to the index, so that the prediction accuracy is improved; generating a causal relationship graph according to the target index, the target feature and the relationship between the target index and the target feature, and fully considering the interrelationship between each index and the factors by carrying out hierarchical division on the target index and the influencing factors, thereby improving the prediction accuracy; according to the causal relationship graph, a dynamic model for predicting power supply safety is established, the superiority of the dynamic model in dynamic data prediction and long-term prediction is utilized, and the prediction accuracy is improved; according to the dynamic model, the power supply safety condition of the snowflake type power distribution network in a future preset time period is predicted, the technical problem that the prediction accuracy of the power supply safety condition of the snowflake type power distribution network in the prior art is low is solved, and the technical effect of accurately predicting the power supply safety condition of the snowflake type power distribution network is achieved.

In an alternative embodiment, determining at least one target indicator for a snowflake-type power distribution network includes:

It should be noted that, the initial index may be obtained by an expert method, the action mechanism of the expert method is to collect the professional knowledge related to the field, provide the professional opinion and knowledge by means of questionnaire, interview, group discussion and the like, answer the research questions, integrate and analyze the information provided by the expert to form qualitative conclusions or suggestions, obtain the initial index, then construct a fishbone graph for obtaining the target index from the initial index, and determine the target index according to the fishbone graph.

In an alternative embodiment, as shown in FIG. 2: fig. 2 is a schematic diagram of a fishbone graph combing initial index hierarchy, wherein specific indexes to be predicted are determined according to actual requirements, the dimensions affecting snowflake-type power distribution network indexes are numerous, the fishbone graph of the affected dimensions is drawn by combining a literature arrangement method and an expert discussion method, then the fishbone graph is combed according to the action relation hierarchy affecting the dimensions, and the dimensions affecting the indexes are summarized by drawing the fishbone graph by adopting the literature arrangement and the expert discussion method.

In an alternative embodiment, determining at least one target feature corresponding to the target indicator includes:

generating a first matrix according to the target index;

and determining target characteristics according to the VIP score.

In an alternative embodiment, as shown in FIG. 3: fig. 3 is a schematic diagram of determining target characteristics by PLS-VIP method according to an embodiment of the present invention, which specifically includes: and normalizing the data by using a maximum and minimum method according to the normalized matrices and/> , and establishing an influence factor matrix/> and a target variable matrix/> . Determining the number of potential variables through an explanatory test and an LOO cross test, establishing an initial model to contain all influencing factors, establishing a partial least squares regression model, acquiring the relation between a response variable and an explanatory variable, establishing the potential variable by maximizing the covariance of the explanatory variable and the response variable, and each principal component is associated with a weight vector between the response variable and the explanatory variable, wherein the weight vector is used for projecting the original explanatory variable into a principal component space. Calculating VIP values of the variables, and measuring importance of the explanatory variables by calculating contribution degree of each explanatory variable to the main component; an interpretation variable with a high VIP value contributes significantly to the model, whereas it contributes little. An explanatory test is developed to evaluate the extent of contribution of each potential variable to the total variance, and a threshold value for the explanatory test is selected based on the variance ratio. And gradually selecting potential variables with VIP values larger than a set threshold value from the initial model according to the degree of contribution degree of the potential variables to the total variance, and calculating the total variance contribution of the selected potential variables. Each subsequent step adds the next most important latent variable to the model, recalculates the total variance contribution ratio of the selected latent variable, and stops selecting the latent variable when the selected latent variable total variance contribution ratio exceeds a predetermined explanatory check threshold. Constructing a model using the selected latent variables and evaluating the performance of the model using LOO cross-checking; each time a LOO cross-check is performed, one latent variable is removed from the dataset and then model fitting is performed using the remaining latent variables; the performance metric may be measured using root mean square error by predicting the removed latent variables using the fitted model and calculating the prediction error. The above steps are repeated to calculate LOO cross-check scores representing the average performance of the model, and the number of potential variables is selected based on the cross-check scores to minimize the LOO-CV score. The model with the best predicted performance can be obtained by selecting the number of potential variables with the lowest LOO-CV score. By combining the interpretive test with the LOO cross test, the selection of the appropriate number of potential variables allows the model to both account for sufficient variance and provide good predictive performance. Overfitting is avoided while selecting the number of potential variables with the highest information value. And mapping the screened potential variables into original influence factors, wherein the influence factors are main influence factors and serve as input variables of a system dynamics model.

Based on the above embodiments, the present invention provides an alternative embodiment, as shown in table 1: table 1 provides a plurality of initial indexes affecting the power supply safety of the snowflake type power distribution network, and a plurality of initial characteristics corresponding to each initial index, and the target indexes are obtained through the hierarchical analysis of the fishbone graph. After the target index and the corresponding initial feature are obtained, calculating the VIP value of the initial feature by using the VIP value of the influence factor of PLS-VIP on the index to obtain the target feature.

TABLE 1

As shown in tables 2 and 3: table 2 provides VIP values for a plurality of initial features corresponding to target metrics calculated using the PLS-VIP method. Table 3 provides the target features screened based on VIP values of the initial features.

TABLE 2

Sequence number	Influencing factors	VIP value	Sequence number	Influencing factors	VIP value
						1	Net rack investment (containing ratio)	1.3208	13	Population speed increasing (containing ratio)	1.4947
2	110KV transformation capacity (capacitance-to-load ratio)	1.3112	14	Population growth rate (containing ratio)	1.3874
						3	35KV transformation capacity (capacitance-to-load ratio)	1.5307	15	GDP (volume ratio)	1.3603
4	10KV transformation capacity (capacitance-to-load ratio)	1.2312	16	GDP speed increasing (capacity ratio)	1.3409
						5	110KV capacity ratio (capacity ratio)	1.3975	17	GDP growth Rate (Capacity ratio)	1.3232
6	35KV capacity-to-load ratio (capacity-to-load ratio)	1.5575	1	Grid investment (load transfer rate)	1.5009
						7	10KV capacity-to-load ratio (capacity-to-load ratio)	1.6521	2	Perfecting net frame type investment ratio (load transfer rate)	1.4762
8	110KV power supply load (capacity-load ratio)	1.2293	3	10KV meets the N-1 main transformer duty ratio (load transfer rate)	1.5015
						9	35KV power supply load (capacity-load ratio)	1.4161	4	10KV meets the N-1 line duty cycle (load transfer rate)	1.0569
10	10KV power supply load (capacity-load ratio)	1.3939	5	35KV meets the N-1 main transformer duty ratio (load transfer rate)	1.3328
						11	Full society maximum load (capacity-to-load ratio)	1.2898	6	35KV meets the N-1 line duty cycle (load transfer rate)	1.3871
12	Power supply population (containing ratio)	1.4961	7	110KV meets the N-1 main transformer duty ratio (load transfer rate)	1.2026
									8	110KV meets the N-1 line duty ratio (load transfer rate)	1.2201

TABLE 3 Table 3

In an alternative embodiment, as shown in fig. 4 and 5: fig. 4 is a schematic diagram of a relationship between a load transfer rate and a corresponding target feature provided by an embodiment of the present invention, and fig. 5 is a schematic diagram of a relationship between a capacity/load ratio and a corresponding target feature provided by an embodiment of the present invention. Dividing the main influencing factors of the prediction index into different layers can make the hierarchical structure clear and the complex interrelationship among the main influencing factors convenient to clear. The system dynamics simulation model is used for describing the state development and change process of the snowflake type power distribution network in a certain area, clearing the logic relationship among main influence factors of each index, and accurately and reasonably designing the variable description hierarchical structure relationship. For the influence factors with unclear hierarchical action relationship, combing can be assisted by combining other methods so as to simplify the dynamic modeling process. After modeling, a system dynamics relation equation among influence factors is established through proper simplification and quantization processing, and the dynamic development process of the regional snowflake type power distribution network system is simulated and the future index condition of the regional snowflake type power distribution network system is predicted.

In an alternative embodiment, generating a causal relationship graph based on target metrics, target characteristics, and relationships between target metrics and target characteristics, includes:

In an alternative embodiment, as shown in fig. 6, fig. 6 is a schematic diagram of a causal relationship graph provided in an embodiment of the present invention, in which a level of an action relationship between a target index and a target feature is represented in the graph in a node form, an arrow indicates a causal relationship, the arrow points in a direction of indicating the causal relationship, and then whether the causal relationship is positive or negative is determined according to service logic, so as to find a feedback loop and a delay structure in the system graph, and perform improvement and beautification on the basis of the feedback loop and the delay structure.

In an alternative embodiment, building a kinetic model for predicting power safety from a causal relationship graph, comprises:

Generating a stack flow graph according to the causal relation graph;

In an alternative embodiment, as shown in fig. 7, fig. 7 is a schematic diagram of a stack flow graph provided in an embodiment of the present invention, where the quantity in the system can be quantified and measured conveniently by converting a causal relationship graph into a stack flow graph, and the variables and data input system dynamics software can be simulated and analyzed conveniently, and the inflow and outflow in the stack flow graph represent the direction and process of influencing accumulation. And selecting a power supply population, GDP and GDP growth rate from the causal relation graph as starting points to establish a stack flow graph, determining the direction of information flow according to the arrow direction in the causal relation graph, forming an information flow path, and gradually adding other nodes according to the arrow direction until the information flow path covering the whole causal relation graph.

Based on the above embodiments, the present invention also provides an alternative embodiment, as shown in table 4: table 4 provides the types of each node in the stack flow graph in the dynamics model, including a (auxliary), R (rate), E (explicit), etc.

TABLE 4 Table 4

In the embodiment of the invention, the load transfer rate and the capacity ratio are horizontal variables, the establishment of a model equation mainly refers to historical statistical data, the correlation between variables is found by utilizing SPSS regression analysis, a certain area of TJ city is taken as a verification area, model input data is input according to the historical data of the verification area 2010-2020, and a variable relation equation is input by using a visual operation interface of VENSIM PLE software.

In an alternative embodiment, the model is interpreted and adjusted based on the validation results, indicating that the model validation is more successful if the model fits better to the actual observed data and reasonable conclusions are drawn through sensitivity analysis. If the model simulation result is greatly different from the actual observed data, the model needs to be adjusted and optimized according to the verification result, for example, parameter values may need to be estimated again, and initial conditions or structural equations may be adjusted to improve the fitting and prediction capability of the model. In particular, when the application software performs model verification, the verification of the model is mainly divided into two parts, namely visual logic verification and operation result verification, and the verification of the operation result is focused. Visual logic verification is that whether the target and the dividing boundary of the system are reasonable or not is checked visually, whether the logic causal relationship between variables is clear or not is checked visually, and whether a model equation is reasonable and correct or not is checked visually; and secondly, running the model on a computer through software, and performing equation checking and dimension checking to ensure the accuracy of the equation expression of the simulation model. The running result verification is to verify the consistency of the behavior of the model and the actual system, and mainly adopts a statistical method to verify the simulation result of the system dynamics model. The running result of the model is generally compared with the historical data to simulate the degree of deviation of the data from the historical data, and whether the model behavior is consistent with the actual system is checked. The relative deviation formula is: Wherein is the relative degree of deviation of/> years/> factor simulation results; the/> is the simulation operation result of the/> years/> factors, the history data of the/> years/> factors, the variable number of the/> is generally more than 70% and the relative deviation is not more than 10%, and the model is considered to perform well, otherwise, the model and related parameters need to be modified for re-verification.

In an alternative embodiment, predicting the power supply safety condition of the snowflake type power distribution network in a future preset time period according to a dynamics model includes:

Specifically, the power supply safety condition of the snowflake type power distribution network in a preset time period in the future can be predicted according to the relation between the capacity-to-load ratio and the power supply safety of the snowflake type power distribution network and the relation between the load transfer rate and the power supply safety of the snowflake type power distribution network, wherein the relation between the capacity-to-load ratio and the power supply safety of the snowflake type power distribution network and the relation between the load transfer rate and the power supply safety of the snowflake type power distribution network can be obtained in advance by related personnel.

In an alternative embodiment, taking a certain area in TJ city as a verification area, taking a power distribution system in 2010-2025 as an example of the verification area for historical simulation, the main output indexes are load transfer rate and capacity ratio, the initial value of GDP is set to 14013.4 hundred million yuan, the population is set to 6589 ten thousand people, the initial value of regional power grid investment is 19.73 hundred million yuan, as shown in table 5 and table 6, table 5 provides the comparison result of the load transfer rate predicted value and the actual value in 2010-2020 of the verification area, and table 6 provides the comparison result of the capacity ratio predicted value and the actual value in 2010-2020 of the verification area.

TABLE 5

Time of	Load transfer rate simulation value	Actual value of load transfer rate	Error value	Relative error (%)
					2010(1)	57.69	60.06	-2.37	-3.95
2010(2)	59.15	57.78	1.37	2.37
					2010(3)	59.32	61.45	-2.13	-3.47
2010(4)	60.22	61.54	-1.32	-2.14
					2011(1)	60.98	63.56	-2.58	-4.06
2011(2)	61.84	63.06	-1.22	-1.93
					2011(3)	62.52	65.12	-2.6	-3.99
2011(4)	61.78	64.14	-2.36	-3.68
					2012(1)	59.72	65.09	-5.37	-8.25
2012(2)	63.66	65.23	-1.57	-2.41
					2012(3)	64.76	68.13	-3.37	-4.95
2012(4)	65.38	67.36	-1.98	-2.94
					2013(1)	65.32	68.98	-3.66	-5.31
2013(2)	65.77	68.23	-2.46	-3.61
					2013(3)	65.34	69.99	-4.65	-6.64
2013(4)	66.38	69.61	-3.23	-4.64
					2014(1)	66.11	70.38	-4.27	-6.07
2014(2)	66.06	71.54	-5.48	-7.66
					2014(3)	66.17	71.97	-5.8	-8.06
2014(4)	66.25	72.46	-6.21	-8.57
					2015(1)	65.01	72.34	-7.33	-10.14
2015(2)	66.26	71.38	-5.12	-7.17
					2015(3)	65.34	73.01	-7.67	-10.51
2015(4)	64.9	73.02	-8.12	-11.12
					2016(1)	66.05	72.52	-6.47	-8.92
2016(2)	67.01	73.05	-6.04	-8.27
					2016(3)	66.44	72.88	-6.44	-8.84
2016(4)	66.42	73.02	-6.6	-9.04
					2017(1)	66.44	74.01	-7.57	-10.23
2017(2)	66.32	74.18	-7.86	-10.59
					2017(3)	66.47	73.23	-6.76	-9.23
2017(4)	65.52	75.19	-9.67	-12.86
					2018(1)	66.73	71.89	-5.16	-7.178
2018(2)	65.5	71.35	-5.85	-8.2
					2018(3)	65.09	72.03	-6.94	-9.63
2018(4)	65.98	72.92	-6.94	-9.52
					2019(1)	62.74	67.75	-5.01	-7.39
2019(2)	63.76	68.34	-4.58	-6.7
					2019(3)	63.28	67.97	-4.69	-6.9
2019(4)	63.31	69.21	-5.9	-8.52
					2020(1)	64.54	70.33	-5.79	-8.23
2020(2)	65.22	71.22	-6	-8.42
					2020(3)	63.78	72.01	-8.23	-11.43
2020(4)	63.59	71.87	-8.28	-11.52

TABLE 6

Time of	Load-to-capacitance ratio simulation	Actual capacity ratio	Error value	Relative error (%)
					2010(1)	1.74	1.73	0.01	0.58
2010(2)	1.71	1.75	-0.04	-2.29
					2010(3)	1.72	1.78	-0.06	-3.37
2010(4)	1.67	1.72	-0.05	-2.91
					2011(1)	1.63	1.74	-0.11	-6.32
2011(2)	1.62	1.74	-0.12	-6.89
					2011(3)	1.64	1.77	-0.13	-7.34
2011(4)	1.63	1.76	-0.13	-7.39
					2012(1)	1.59	1.71	-0.12	-7.02
2012(2)	1.58	1.7	-0.12	-7.06
					2012(3)	1.54	1.7	-0.16	-9.41
2012(4)	1.54	1.68	-0.14	-8.33
					2013(1)	1.59	1.67	-0.08	-4.79
2013(2)	1.54	1.63	-0.09	-5.52
					2013(3)	1.55	1.63	-0.08	-4.91
2013(4)	1.54	1.64	-0.1	-6.1
					2014(1)	1.55	1.72	-0.17	-9.88
2014(2)	1.55	1.72	-0.17	-9.88
					2014(3)	1.56	1.73	-0.17	-9.83
2014(4)	1.57	1.73	-0.16	-9.25
					2015(1)	1.58	1.71	-0.13	-7.6
2015(2)	1.6	1.7	-0.1	-5.88
					2015(3)	1.61	1.72	-0.11	-6.4
2015(4)	1.61	1.72	-0.11	-6.39
					2016(1)	1.63	1.71	-0.08	-4.68
2016(2)	1.62	1.72	-0.1	-5.81
					2016(3)	1.62	1.74	-0.12	-6.9
2016(4)	1.64	1.73	-0.09	-5.2
					2017(1)	1.63	1.7	-0.07	-4.12
2017(2)	1.61	1.69	-0.08	-4.73
					2017(3)	1.6	1.72	-0.12	-6.98
2017(4)	1.58	1.66	-0.08	-4.82
					2018(1)	1.57	1.7	-0.13	-7.65
2018(2)	1.58	1.71	-0.13	-7.6
					2018(3)	1.58	1.7	-0.12	-7.06
2018(4)	1.6	1.68	-0.08	-4.76
					2019(1)	1.6	1.71	-0.11	-6.43
2019(2)	1.64	1.72	-0.08	-4.65
					2019(3)	1.64	1.73	-0.09	-5.2
2019(4)	1.65	1.74	-0.09	-5.17
					2020(1)	1.66	1.74	-0.08	-4.6
2020(2)	1.7	1.76	-0.06	-3.41
					2020(3)	1.73	1.78	-0.05	-2.81
2020(4)	1.76	1.79	-0.03	-1.68

As shown in fig. 8, fig. 8 is a graph of load transfer rate predicted values versus actual values in the verification areas 2010-2020 according to an embodiment of the present invention. As shown in Table 7, table 7 provides predicted values and deviation magnitudes of target indexes, and the relative deviation of the predicted results is 3%, which indicates that the provided index prediction method has good prediction accuracy.

TABLE 7

Index (I)	Actual statistics in 2020	2020 Simulation predictive value	Deviation amplitude	Relative deviation of
					Load transfer rate	72.69	72.11	-0.58	-0.80%
Ratio of capacity to load	1.82	1.81	-0.01	-0.55%

It should be noted that a prominent advantage of system dynamics is that it can cope with the problems of high-order, nonlinear and complex time-varying systems, and can decompose the mechanism of the system into various simple entities and elements, so that the complex phenomenon existing in the system can be disassembled into a simple and effective mechanism, and multiple analyses can be conveniently performed on the system. Through analysis of the model causal relationship graph and reasonable selection of model equations and variables, the prediction accuracy of results is greatly improved, and accurate prediction of the actual state of the future year of the snowflake type power distribution network is facilitated.

The method for combining the dynamic model and the snowflake type power distribution network index prediction provided by the embodiment of the invention has the following application prospect:

(1) Long-term planning and policy formulation: the system can be used for evaluating the influence of different planning schemes on the capacity-to-load ratio and the load transfer rate by establishing a system dynamics model, so as to formulate sustainable development and capacity planning strategies. And a regulation strategy for optimizing the capacity-to-load ratio or the load transfer rate can be obtained by adjusting factors such as the 10kV capacity-to-load ratio, the 35kV transformation capacity, the investment scale and the like.

(2) Risk management: the system dynamics model can be used for analyzing the risks of the capacity ratio and the load transfer rate under different conditions, helping a power grid operator to make a risk management strategy, optimizing a capacity planning decision, upgrading or expanding equipment in time, and preventing overload and unstable conditions.

(3) Capacity planning and extension: the system dynamics can help planners evaluate future load demands, predict whether the system capacity is sufficient and whether equipment needs to be upgraded or expanded, help optimize capacity planning decisions, and ensure the reliability and stability of the power grid.

(4) Device status monitoring and maintenance: by establishing a system dynamics model, the states of various devices in the snowflake type power distribution network, such as a distribution transformer, a pole-mounted switch, a ring network unit, a circuit and the like, can be monitored, the service life of the devices and the maintenance requirements are predicted, and the utilization rate and the reliability of the devices are improved.

(5) And (3) electric energy quality analysis: system dynamics can be used to analyze power quality problems to determine the root cause of the problem and provide a solution.

(6) Electric power market simulation: in power market simulation, system dynamics methods may be used to simulate market participant behavior, pricing mechanisms, and trading rules to assess market efficiency and competitiveness.

Based on the same conception, the embodiment of the invention provides a snowflake type power distribution network power supply safety prediction device, which is shown in fig. 9: fig. 9 is a schematic structural diagram of a snowflake type power distribution network power supply safety prediction device according to an embodiment of the present invention, where the device includes:

The determining module 201 is configured to determine at least one target index of the snowflake type power distribution network, and at least one target feature corresponding to the target index, where the target index includes an index affecting power supply safety of the snowflake type power distribution network, and the target feature includes a power factor affecting the target index;

a generating module 202, configured to generate a causal relationship graph according to the target index, the target feature, and the relationship between the target index and the target feature;

The establishing module 203 is configured to establish a dynamics model for predicting power supply safety according to the causal relationship graph;

And the prediction module 204 is used for predicting the power supply safety condition of the snowflake type power distribution network in a preset time period in the future according to the dynamics model.

Further, the determining module 201 includes an acquiring unit and a first determining unit;

Further, the determining module 201 further includes a first generating unit, a second generating unit, a third generating unit, a calculating unit, and a second determining unit;

Further, the generating module 202 includes a node determining unit and an edge generating unit;

Further, the setup module 203 includes a fourth generation unit, a first setup unit, and a second setup unit;

Further, the prediction module 204 includes a first prediction unit and a second prediction unit;

The apparatus embodiments described above are merely illustrative, wherein elements illustrated as separate elements may or may not be physically separate, and elements shown as elements may or may not be physical elements, may be located in one place, or may be distributed over a plurality of network elements. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

As shown in fig. 10, the embodiment of the present invention further provides an apparatus, which includes a processor 301, a communication interface 302, a memory 303, and a communication bus 304, where the processor 301, the communication interface 302, and the memory 303 perform communication with each other through the communication bus 304.

A memory 303 for storing a computer program.

In one embodiment of the present invention, when executing a program stored in the memory 303, the processor 301 is configured to implement the snowflake power distribution network power supply security prediction method provided in any one of the foregoing method embodiments, where the method includes:

The embodiment of the invention also provides a computer readable storage medium, on which a computer program is stored, which when executed by a processor, implements the snowflake type power distribution network power supply safety prediction method provided by any one of the method embodiments, comprising:

From the above description of embodiments, it will be apparent to those skilled in the art that the embodiments may be implemented by means of software plus a general purpose hardware platform, or may be implemented by hardware. Based on such understanding, the foregoing technical solution may be embodied essentially or in a part contributing to the related art in the form of a software product, which may be stored in a computer readable storage medium, such as ROM/RAM, a magnetic disk, an optical disk, etc., including several instructions to cause a computer device (which may be a personal computer, a server, or a network device, etc.) to execute the method of the respective embodiments or some parts of the embodiments.

It is to be understood that the terminology used herein is for the purpose of describing particular example embodiments only, and is not intended to be limiting. As used herein, the singular forms "a", "an" and "the" are intended to include the plural forms as well, unless the context clearly indicates otherwise. The terms "comprises," "comprising," "includes," "including," and "having" are inclusive and therefore specify the presence of stated features, steps, operations, elements, and/or components, but do not preclude the presence or addition of one or more other features, steps, operations, elements, components, and/or groups thereof. The method steps, processes, and operations described herein are not to be construed as necessarily requiring their performance in the particular order described or illustrated, unless an order of performance is explicitly stated. It should also be appreciated that additional or alternative steps may be used.

The foregoing is merely exemplary of embodiments of the present invention to enable those skilled in the art to understand or practice the invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the generic principles defined herein may be applied to other embodiments without departing from the spirit or scope of the invention. Thus, the present invention is not intended to be limited to the embodiments shown herein but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A snowflake type power distribution network power supply safety prediction method is characterized by comprising the following steps:

determining at least one target index of the snowflake type power distribution network and at least one target feature corresponding to the target index, wherein the target index comprises an index affecting the power supply safety of the snowflake type power distribution network, and the target feature comprises a power factor affecting the target index;

the determining at least one target index of the snowflake type power distribution network comprises the following steps:

Acquiring at least one initial index affecting the power supply safety, and generating a fishbone graph according to the initial index, wherein the fishbone graph is used for combing the layers of the initial index;

Determining the target index according to the fishbone graph, wherein the target index comprises a capacity-to-load ratio and a load transfer rate;

the determining at least one target feature corresponding to the target index includes:

generating a first matrix according to the target index;

Generating a second matrix according to at least one initial feature affecting the target index and according to the initial feature;

generating a PLS regression model from the first matrix and the second matrix, comprising:

establishing a partial least squares regression model containing all target features, acquiring a relation between an initial feature and an initial index, and creating a potential variable by maximizing covariance between the initial feature and the initial index, wherein each principal component is associated with a weight vector between the initial index and the initial feature, and the weight vector is used for projecting the initial feature into a principal component space;

calculating a VIP score for each of the initial features according to the PLS regression model;

determining the target feature according to the VIP score, including:

developing an explanatory test to evaluate the contribution of each of the potential variables to the total variance, selecting a threshold for the explanatory test based on the variance ratio, selecting the potential variables with VIP values greater than a set threshold from the partial least squares regression model, calculating the total variance contribution of the selected potential variables, gradually adding the next most significant potential variable to the model, and calculating the total variance contribution ratio of the selected potential variables, stopping selecting the potential variables when the total variance contribution ratio of the selected potential variables exceeds a predetermined explanatory test threshold, constructing a model using the selected potential variables, and evaluating the performance of the model using a lot of lot cross tests, removing one of the potential variables from a dataset when the lot of cross tests are performed, using the model of the rest, predicting the removed potential variables using the fitted model, and calculating a prediction error, using a root mean square error, repeatedly performing a performance metric, using the lot of cross tests, selecting the potential variables as the number of cross tests, using the number of the cross tests, selecting the potential variables as the threshold, using the lot of cross tests, and using the lot of potential variables as the rest scores;

Generating a causal relationship graph according to the target index, the target feature, and the relationship between the target index and the target feature, including:

taking the target index and the target feature as nodes, wherein the nodes comprise affected nodes and affected nodes;

connecting the nodes with the relation by using the edges with the arrows to obtain the causal relation graph, wherein the directions of the arrows point to the affected factors from the affected factors;

Establishing a dynamics model for predicting the power supply safety according to the causal relation graph, wherein the dynamics model comprises the following steps of:

generating a stack flow graph according to the causal relation graph;

According to the stack flow graph, a first sub-model for predicting the capacity ratio is established, wherein the first sub-model comprises a first state variable equation, a first rate equation and a first auxiliary equation, and the first state variable equation is: ,/> Is the value of the load transfer rate/> moment,/> is the value of the load transfer rate initial moment/> ,/> is the rate of change of the load transfer rate at the moment t,/> is the input rate of the load transfer rate,/> is the output rate of the load transfer rate, and the first rate equation is: and (3) wherein/> ,/> is an auxiliary variable value of the load transfer rate, wherein/> is an exogenous variable value of the load transfer rate, wherein/> is a linear function, and wherein the first auxiliary equation is as follows: the/> ,/> is an auxiliary variable other than the auxiliary variable of the load transfer rate/> ;

establishing a second sub-model for predicting the load transfer rate according to the stack flow graph, wherein the second sub-model comprises a second state variable equation, a second rate equation and a second auxiliary equation, and the second state variable equation is: ,/> Is the value of the capacity ratio/> , is the value of the capacity ratio at the initial time/> ,/> is the rate at which the capacity ratio changes at time t, is the input rate of the capacity ratio,/> is the output rate of the capacity ratio, and the second rate equation is: and/> ,/> is the value of an auxiliary variable of the capacity-to-capacity ratio, and/> is the value of an exogenous variable of the capacity-to-capacity ratio, the second auxiliary equation being: ,/> Is an auxiliary variable other than the auxiliary variable/> of the loading ratio;

Calculating the relative deviation of the predicted result of the dynamic model according to a deviation formula, wherein the relative deviation formula satisfies the following conditions: ,/> The relative deviation degree of the predicted results of the target indexes of the/> in the/> years, the is the predicted results of the target indexes of the/> in the/> years, the/> is the historical data of the target indexes of the/> in the/> years, when the number of the target indexes of the accounts for more than 70% of all the target indexes, and the relative deviation degree is not more than 10%, the dynamics model is determined to be good, otherwise, the dynamics model is retrained;

according to the dynamics model, predicting the power supply safety condition of the snowflake type power distribution network in a preset time period in the future comprises the following steps:

Predicting the capacity-to-load ratio and the load transfer rate of the snowflake-type power distribution network in a preset time period in the future;

2. A snowflake type power distribution network power supply safety prediction device is characterized by comprising:

The determining module is used for determining at least one target index of the snowflake-type power distribution network and at least one target characteristic corresponding to the target index, wherein the target index comprises an index affecting the power supply safety of the snowflake-type power distribution network, and the target characteristic comprises a power factor affecting the target index;

The determining module comprises an obtaining unit and a first determining unit;

The acquisition unit is used for acquiring at least one initial index affecting the power supply safety and generating a fishbone graph according to the initial index, wherein the fishbone graph is used for combing the layers of the initial index;

the first determining unit is configured to determine the target indicator according to the fishbone graph, where the target indicator includes a load ratio and a load transfer rate;

The determining module further comprises a first generating unit, a second generating unit, a third generating unit, a calculating unit and a second determining unit;

the second generation unit is used for generating a second matrix according to at least one initial characteristic affecting the target index and according to the initial characteristic;

The third generating unit is configured to generate a PLS regression model according to the first matrix and the second matrix, and includes:

the calculating unit is used for calculating the VIP score of each initial characteristic according to the PLS regression model;

The second determining unit is configured to determine the target feature according to the VIP score, and includes:

Developing an explanatory test to evaluate the contribution of each of the potential variables to the total variance, selecting a threshold for the explanatory test based on the variance ratio, selecting the potential variables with VIP values greater than a set threshold from the partial least squares regression model, calculating the total variance contribution of the selected potential variables, gradually adding the next most significant potential variable to the model, and calculating the total variance contribution ratio of the selected potential variables, stopping selecting the potential variables when the total variance contribution ratio of the selected potential variables exceeds a predetermined explanatory test threshold, constructing a model using the selected potential variables, and evaluating the performance of the model using a lot of lot cross tests, removing one of the potential variables from a dataset when the lot of cross tests are performed, using the remaining potential variables to model, predicting the removed potential variables using the fitted model, and calculating a prediction error, using a root mean square error to perform a performance metric, repeatedly using the lot of cross tests, removing the selected potential variables from the dataset, using the lot of cross tests as a potential variables, selecting a number of potential variables, using the lot of cross tests as a threshold, and using the lot of cross tests as the potential variables; the generation module is used for generating a causal relationship graph according to the target index, the target feature and the relationship between the target index and the target feature;

The generating module comprises a node determining unit and an edge generating unit;

The node determining unit is used for taking the target index and the target characteristic as nodes, wherein the nodes comprise affected nodes and affected nodes;

the edge generation unit is used for connecting the nodes with the arrow to obtain the causal relationship graph, wherein the direction of the arrow points to the affected factors from the affected factors;

The establishing module is used for establishing a dynamics model for predicting the power supply safety according to the causal relation graph;

the building module comprises a fourth generating unit, a first building unit and a second building unit;

the fourth generating unit is used for generating a stack flow graph according to the causal relation graph;

The first establishing unit is configured to establish a first sub-model for predicting the capacity ratio according to the stack flow graph, where the first sub-model includes a first state variable equation, a first rate equation, and a first auxiliary equation, and the first state variable equation is: ,/> Is the value of the load transfer rate/> moment,/> is the value of the load transfer rate initial moment/> ,/> is the rate of change of the load transfer rate at the moment t,/> is the input rate of the load transfer rate,/> is the output rate of the load transfer rate, and the first rate equation is: and (3) wherein/> , is an auxiliary variable value of the load transfer rate, wherein/> is an exogenous variable value of the load transfer rate, wherein/> is a linear function, and wherein the first auxiliary equation is as follows: the/> ,/> is an auxiliary variable other than the auxiliary variable of the load transfer rate/> ;

the second building unit is configured to build a second sub-model for predicting the load transfer rate according to the stack flow graph, where the second sub-model includes a second state variable equation, a second rate equation, and a second auxiliary equation, and the second state variable equation is: ,/> Is the value of the capacity ratio/> moment,/> is the value of the capacity ratio at initial moment/> ,/> is the rate at which the capacity ratio changes at time t,/> is the input rate of the capacity ratio,/> is the output rate of the capacity ratio, and the second rate equation is: and/> ,/> is the value of an auxiliary variable of the capacity-to-capacity ratio, and/> is the value of an exogenous variable of the capacity-to-capacity ratio, the second auxiliary equation being: ,/> Is an auxiliary variable other than the auxiliary variable/> of the loading ratio;

The prediction module is used for predicting the power supply safety condition of the snowflake type power distribution network in a preset time period in the future according to the dynamics model;

the prediction module comprises a first prediction unit and a second prediction unit;

The first prediction unit is used for predicting the capacity-to-load ratio and the load transfer rate of the snowflake-shaped power distribution network in a preset time period in the future;

3. An electronic device, comprising:

one or more processors;

A memory for storing one or more programs;

the one or more programs, when executed by the one or more processors, cause the one or more processors to implement the snowflake-type power distribution network power supply safety prediction method of claim 1.

4. A computer readable storage medium having stored thereon a computer program which when executed by a processor implements the snowflake power distribution network power security prediction method of claim 1.