Variable grid group early warning method and system oriented to cloud edge collaboration

Abstract

The invention discloses a variable grid group early warning method and system oriented to cloud edge collaboration, and relates to the technical field of variable grid group early warning, and the method comprises the steps: obtaining the data of climate factors, and constructing the characteristic functions of different climate factors about the air flow field of a wind field; constructing a renewable energy output scene, and performing scene generation and scene reduction; the renewable energy source output scene is optimized; constructing a prediction model based on the optimized renewable energy output scene and performing prediction to obtain a power generation trend prediction result; and risk early warning is carried out based on the power generation trend prediction result, and variable grid group early warning oriented to cloud edge collaboration is realized. For the most serious fault scene under the fault uncertainty set, the renewable energy power supply is configured in the planning stage, and the elasticity of the power distribution network is improved; by considering the influence of renewable energy output uncertainty, a renewable energy output scene is constructed, optimization is carried out by taking the minimum load shedding expectation in each scene as a target, and the prediction robustness is improved.

Description

Technical Field

[0001] The present invention relates to the field of variable grid group early warning technology, and in particular to a variable grid group early warning method and system for cloud-edge collaboration. Background Art

[0002] With the rapid development and widespread application of renewable energy, the resource utilization mode of distribution network under the cloud-edge collaborative architecture has promoted the rapid development of multi-link technologies of source-grid-load-storage at the end of energy consumption. At present, the provinces, cities and counties of China Southern Power Grid Company have built platforms at various levels, and paid attention to the interconnection and interoperability of user energy consumption data collection and relevant platform data; however, with the increase in the proportion of distributed power supply access and the increase in the demand for interaction between multiple subjects, the global unified distributed power supply control and management method in the traditional distribution network platform has begun to be restricted; first of all, the existing global unified distributed power supply control and management method can no longer adapt to the increasingly complex power grid structure, and a large number of distributed power supplies are multi-faceted. The wide access mode has brought huge challenges to system dispatching, resulting in increasing difficulty for dispatchers to make decisions; secondly, the communication link between distributed power sources and the system dispatching end is often weak, making it difficult for dispatching commands to arrive in time, resulting in increased difficulty in monitoring. In this case, distributed power sources are prone to disorderly power generation, which not only fails to give play to the advantages of interaction between multiple subjects, but may also increase the pressure on grid operation; in addition, the objects of grid regulation have grown explosively, and the OS2 system designed for the integrated operation of limited master stations and plant stations has encountered bottlenecks in its architecture. The emergence of "separation of transmission and distribution" and integrated energy service providers has gradually promoted the development of new power systems in the direction of regionalization and gridding.

[0003] In response to the relevant needs of renewable energy, the shortcomings of the current common solutions include: lack of comprehensive consideration of climate factors, and the inability to accurately predict the output characteristics of renewable energy; the scenario generation and optimization methods are single, and it is difficult to cope with the complex and changeable power grid operation environment; the prediction model is not accurate enough to provide reliable support for scheduling decisions, which seriously restricts the safe and stable operation of the distribution network and the efficient use of renewable energy; in the context of multi-energy complementarity and integrated energy services, existing technologies are still difficult to achieve coordinated optimization and unified scheduling of multiple energy forms, which restricts the overall efficiency improvement of the energy system. Summary of the invention

[0004] In view of the above-mentioned problems, the present invention is proposed.

[0005] Therefore, the problem to be solved by the present invention is how to provide a variable grid group early warning method for cloud-edge collaboration, by considering the impact of renewable energy output uncertainty, constructing a series of renewable energy output scenarios, and optimizing the configuration with the goal of minimizing the expected load shedding in each scenario, thereby improving the prediction robustness.

[0006] In order to solve the above technical problems, the present invention provides the following technical solutions:

[0007] In the first aspect, an embodiment of the present invention provides a variable grid group early warning method for cloud-edge collaboration, which includes acquiring data on climate factors and constructing characteristic functions of different climate factors on wind field airflow fields; constructing renewable energy output scenarios and performing scenario generation and scenario reduction; optimizing the renewable energy output scenarios; constructing a prediction model based on the optimized renewable energy output scenarios and performing prediction to obtain power generation trend prediction results; and performing risk warnings based on the power generation trend prediction results to achieve variable grid group early warnings for cloud-edge collaboration.

[0008] As a preferred scheme of the variable grid group early warning method for cloud-edge collaboration described in the present invention, wherein: the data of the climate factors is composed of a data set of climate factors corresponding to complex terrain and landforms; the construction of characteristic functions of different climate factors on wind field airflow fields includes the following steps: obtaining the data of climate factors; constructing a mapping relationship between climate factors and wind field airflow fields based on the data of climate factors; constructing characteristic functions of different climate factors on wind field airflow fields based on the mapping relationship between climate factors and wind field airflow fields.

[0009] As a preferred solution of the variable grid group early warning method for cloud-edge collaboration described in the present invention, the construction of renewable energy output scenarios includes: targeting wind speed uncertainty, using Weibull distribution to characterize the airflow field distribution of the wind field; describing the relationship between wind speed and output power; the scenario generation refers to stratified sampling of wind speed using Latin hypercube sampling to generate wind speed sequence scenarios; the scenario reduction refers to determining a subset of an initial scenario and reallocating probabilities for the subset to obtain an allocated subset, so that the probability distance between the probability distribution Q of the allocated subset and the probability distribution P of the subset of the initial scenario is the shortest.

[0010] As an optimal solution of the variable grid group early warning method for cloud-edge collaboration described in the present invention, the optimization of the renewable energy output scenario includes: optimizing the configuration of the renewable energy output scenario with the goal of minimizing the expected load shedding in each scenario; assuming that after the line is attacked, it will be in a fault disconnected state within the optimization time, and then formulating attack and line state constraints, and using the attack and line state constraints to constrain the renewable energy output scenario.

[0011] As a preferred solution of the variable grid group early warning method for cloud-edge collaboration described in the present invention, wherein: the construction of the prediction model and the prediction to obtain the power generation trend prediction result include the following steps: based on the adjacent data extraction method, the original series is obtained and the accumulation operation is performed to obtain the accumulated series; based on the accumulated series, the corresponding differential formula is constructed; based on the discrete formula, the derivative definition, the approximate formula, the equal time interval method and the average evaluation, the original series and the accumulated series are calculated and transformed by identity; the formula obtained after the identity transformation is converted into a matrix expression and calculated to obtain the corresponding vector; based on the principle of solving equations and vectors of predicted values A time response function is constructed, and a cumulative subtraction operation is performed to restore a grey prediction model, i.e., a prediction model; the risk warning comprises the following steps: setting a risk threshold, and using Sigmoid as an activation function to control whether a fault warning is triggered; using the output signal of the activation function as an enable signal for the warning trigger, and using the adjacent value of the risk threshold as a dead zone; mapping the risk assessment value to the comprehensive assessment result for risk warning, realizing fault warning through a trigger signal, and promptly initiating safety precautions.

[0012] As a preferred solution of the variable grid group early warning method for cloud-edge collaboration of the present invention, the calculation formula of the wind field airflow distribution is as follows:

[0013]

[0014] Among them, f w (v) is the probability density function representing the wind speed; k is the shape parameter; c is the scale parameter; v is the wind speed; the calculation formula for the relationship between the wind speed and the output power is as follows:

[0015]

[0016] Among them, P w is the output power; v is the wind speed; v ci is the cut-in wind speed; v rate is the rated wind speed; v co is the cut-out wind speed; P rate is the rated capacity of the wind turbine; the calculation formula of the probability distance between the probability distribution Q of the allocated subset and the probability distribution P of the subset of the initial scenario is as follows:

[0017]

[0018] Among them, D k is the probability distance between the probability distribution Q of the assigned subset and the probability distribution P of the subset of the initial scene; P is the probability distribution of the subset of the initial scene; Q is the probability distribution of the assigned subset; S is the number of scenes in the probability distribution P of the subset of the initial scene; is the number of scenarios of the probability distribution Q of the allocated subset; η ij is the decision variable in the model; For the scene i and scenes The distance between i is a scene in the probability distribution P of a subset of the initial scenes; p i For the scene i Probability of occurrence; is the scenario in the probability distribution Q of the assigned subset; q j For the scene The probability of occurrence; k is the shape parameter.

[0019] As a preferred solution of the variable grid group early warning method for cloud-edge collaboration of the present invention, the calculation formula of the objective function for optimizing the configuration of the renewable energy output scenario is as follows:

[0020]

[0021] Among them, j is a node; V is a node set; h is the cost of configuring distributed renewable energy sources at node j; j is the binary decision variable for whether to configure power supply at node j; h is the decision variable for distributed renewable energy power supply configuration; w is the attack strategy variable vector; z is the distribution network power flow optimization variable vector; r is the continuous variable for topology reconstruction; c is the discrete variable for topology reconstruction; s is the renewable energy output scenario; S is the set of renewable energy output scenarios; ρ s is the probability of renewable energy output scenario s; t is the optimized time period; T is the optimized time period set; is the active power cut-off value of node j in time period t under renewable energy output scenario s; the calculation formula of the attack and line state constraints is as follows:

[0022]

[0023]

[0024] in, is a binary variable representing the open and closed state of the line; u ij is a binary variable indicating whether the line (i, j) is attacked; t is the optimized time period; s is the renewable energy output scenario; U is the attack uncertainty set; L is the combination of distribution network lines; B is the attack budget for the maximum number of attack lines.

[0025] On the second aspect, in order to further solve the security problems existing in the variable grid group early warning, the embodiment of the present invention provides a variable grid group early warning system for cloud-edge collaboration, which includes: a characteristic function module, which is used to obtain the data of corresponding climate factors under complex terrain and construct characteristic functions of different climate factors on wind field airflow fields; a scenario construction module, which is used to construct renewable energy output scenarios and perform scenario generation and scenario reduction according to wind speed uncertainty; a scenario optimization module, which is used to optimize the configuration of renewable energy output scenarios and formulate attack and line status constraints; a trend prediction module, which is used to construct a prediction model based on the optimized renewable energy output scenario and perform prediction to obtain a power generation trend prediction result; a risk warning module, which is used to make judgments based on the power generation trend prediction results and risk thresholds to achieve variable grid group early warning for cloud-edge collaboration.

[0026] In a third aspect, an embodiment of the present invention provides a computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: when the computer program is executed by the processor, any step of a variable grid group early warning method for cloud-edge collaboration as described in the first aspect of the present invention is implemented.

[0027] In a fourth aspect, an embodiment of the present invention provides a computer-readable storage medium having a computer program stored thereon, wherein: when the computer program is executed by a processor, it implements any step of a variable grid group early warning method for cloud-edge collaboration as described in the first aspect of the present invention.

[0028] The beneficial effects of the present invention are as follows: the present invention configures renewable energy sources in the planning stage for the most serious fault scenario under the fault uncertainty set, thereby improving the flexibility of the distribution network; by considering the impact of renewable energy output uncertainty, a series of renewable energy output scenarios are constructed, and the optimization configuration is performed with the goal of minimizing the expected load shedding in each scenario, thereby improving the robustness of the prediction. BRIEF DESCRIPTION OF THE DRAWINGS

[0029] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for describing the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work. Among them:

[0030] Figure 1 This is an overall flow chart of a variable grid group early warning method for cloud-edge collaboration in Example 1.

[0031] Figure 2 This is a flow chart of constructing the characteristic function of the wind field airflow field in Example 1.

[0032] Figure 3 This is a risk warning flow chart of a variable grid group warning method for cloud-edge collaboration in Example 1.

[0033] Figure 4 This is a schematic diagram of the structure of the computer device in Example 3.

[0034] Figure 5 This is a schematic diagram of the CPS62 node calculation example in Example 4.

[0035] Figure 6 This is a typical daily load curve diagram for various areas in Example 4.

[0036] Figure 7 Schematic diagram of the comprehensive risk value of voltage over-limit at nodes in each area in Example 4.

[0037] Figure 8 This is a schematic diagram of the comprehensive risk value of line power overload in each area in Example 4.

[0038] Fig. 9 This is a schematic diagram of the comprehensive risk value of each area in Example 4. DETAILED DESCRIPTION

[0039] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the specific implementation methods of the present invention are described in detail below in conjunction with the accompanying drawings.

[0040] In the following description, many specific details are set forth to facilitate a full understanding of the present invention, but the present invention may also be implemented in other ways different from those described herein, and those skilled in the art may make similar generalizations without violating the connotation of the present invention. Therefore, the present invention is not limited to the specific embodiments disclosed below.

[0041] Secondly, the term "one embodiment" or "embodiment" as used herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The term "in one embodiment" that appears in different places in this specification does not necessarily refer to the same embodiment, nor does it refer to a separate or selective embodiment that is mutually exclusive with other embodiments.

[0042] Example 1

[0043] Reference Figure 1 to Figure 3 , which is the first embodiment of the present invention, and this embodiment provides a variable grid group early warning method for cloud-edge collaboration.

[0044] The existing variable grid group early warning methods mainly have the following problems: first, lack of comprehensive consideration of climate factors, unable to accurately predict the output characteristics of renewable energy; scenario generation and optimization methods are single, difficult to cope with the complex and changeable power grid operation environment; the prediction model is not accurate enough to provide reliable support for scheduling decisions, which seriously restricts the safe and stable operation of the distribution network and the efficient use of renewable energy; in the context of multi-energy complementarity and integrated energy services, existing technologies are still difficult to achieve coordinated optimization and unified scheduling of multiple energy forms, restricting the overall efficiency improvement of the energy system.

[0045] The present application provides an effective solution to the above-mentioned problems. Next, we will explain in detail how to implement the variable grid group warning method for cloud-edge collaboration in combination with multiple embodiments.

[0046] Figure 1 The overall flow chart of a variable grid group early warning method for cloud-edge collaboration is shown, including:

[0047] S1: Obtain the data of climate factors and construct the characteristic functions of different climate factors on wind field, airflow and flow field.

[0048] Preferably, the data of climate factors are composed of data sets of climate factors corresponding to complex topography.

[0049] Furthermore, climate factors include temperature T, wind speed RH, ice cover P, and lightning Q.

[0050] Specifically, Figure 2 The flowchart shown is for constructing the characteristic function of the wind farm airflow flow field. Constructing the characteristic function of the wind farm airflow flow field with different climate factors includes the following steps: obtaining the data of the climate factors.

[0051] Based on the data of climate factors, a mapping relationship between climate factors and wind field, airflow and flow field is constructed.

[0052] Based on the mapping relationship between climate factors and wind farm airflow, the characteristic function of different climate factors on wind farm airflow is constructed. The specific formula is as follows:

[0053] Flow Characteristics=f(T,RH,P,Q,...)

[0054] Among them, Flow Characteristics is a collection of wind field airflow characteristics, including speed, direction and vorticity; f is the mapping function; T is temperature; RH is wind speed; P is icing; and Q is lightning.

[0055] Preferably, by constructing a mapping relationship between climate factors and wind field airflow, a comprehensive and complex characteristic function is constructed, which takes into account multiple climate factors such as temperature, wind speed, icing and lightning, and can more accurately describe the wind field characteristics under complex terrain. Especially under complex terrain conditions, compared with the traditional method that only considers a single factor, the present invention can provide more comprehensive and accurate wind field forecasts; it can capture the interaction between multiple climate factors and better predict wind field behavior; it lays a more reliable data foundation for subsequent scenario generation and optimization, and improves the credibility of the early warning.

[0056] S2: Construct renewable energy output scenarios and perform scenario generation and scenario reduction.

[0057] Preferably, constructing a renewable energy output scenario includes: using Weibull distribution to describe the wind field airflow distribution in view of wind speed uncertainty, and the calculation formula of the wind field airflow distribution is as follows:

[0058]

[0059] Among them, f w (v) is the probability density function representing wind speed; k is the shape parameter; c is the scale parameter; v is the wind speed.

[0060] The relationship between wind speed and output power is described. The calculation formula for the relationship between wind speed and output power is as follows:

[0061]

[0062] Among them, P w is the output power; v is the wind speed; v ci is the cut-in wind speed; v rate is the rated wind speed; v co is the cut-out wind speed; P rate is the rated capacity of the wind turbine.

[0063] Furthermore, scenario generation refers to stratified sampling of wind speed using Latin hypercube sampling to generate wind speed sequence scenarios.

[0064] Furthermore, scene reduction refers to determining a subset of initial scenes and reallocating probabilities for the subset to obtain a subset after allocation, so that the probability distance between the probability distribution Q of the allocated subset and the probability distribution P of the subset of the initial scenes is the shortest.

[0065] Specifically, the calculation formula of the probability distance between the probability distribution Q of the assigned subset and the probability distribution P of the subset of the initial scene is as follows:

[0066]

[0067] Among them, D k is the probability distance between the probability distribution Q of the assigned subset and the probability distribution P of the subset of the initial scene, i.e., Kantorovich distance; P is the probability distribution of the subset of the initial scene; Q is the probability distribution of the assigned subset; S is the number of scenes of the probability distribution P of the subset of the initial scene; is the number of scenarios of the probability distribution Q of the allocated subset; η ij is the decision variable in the model; For the scene i and scenes The distance between them is solved by forward selection and backward selection; ξ i is a scene in the probability distribution P of a subset of the initial scenes; p i For the scene i Probability of occurrence; is the scenario in the probability distribution Q of the assigned subset; q j For the scene The probability of occurrence; k is the shape parameter.

[0068] It should be noted that when constructing renewable energy output scenarios and performing scenario generation and scenario reduction, the methods can also be adjusted based on different needs. For example, based on historical wind power generation data, Monte Carlo simulation can be used to generate multiple possible wind power generation scenarios, and then the K-means clustering method can be used to reduce the scenarios. Alternatively, a solar radiation intensity prediction model can be used to generate multiple photovoltaic power generation scenarios, and then the principal component analysis PCA method can be used to reduce the scenarios and retain the most representative scenarios.

[0069] Preferably, since Weibull distribution can well fit the wind speed distribution characteristics of most areas in wind speed simulation, Weibull distribution is used to characterize the airflow distribution of wind field, and shape parameter k and scale parameter c are introduced to flexibly adapt to the wind speed characteristics of different regions; in terms of scene generation and reduction, Latin hypercube sampling is used for stratified sampling, which has better sample representativeness and computational efficiency than traditional random sampling. In the process of scene reduction, the scene subset is optimized by minimizing the Kantorovich distance to ensure that the reduced scene can still maintain the key characteristics of the original distribution, thereby improving the computational efficiency without losing the prediction accuracy.

[0070] S3: Optimize renewable energy output scenarios.

[0071] Preferably, optimizing the renewable energy output scenario includes: optimizing the configuration of the renewable energy output scenario with the minimum load shedding expectation under each scenario as the goal, improving the prediction robustness, and the calculation formula of the objective function for optimizing the configuration of the renewable energy output scenario is as follows:

[0072]

[0073] Among them, j is a node; V is a node set; h is the cost of configuring distributed renewable energy sources at node j; j is the binary decision variable for whether to configure power supply at node j; h is the decision variable for distributed renewable energy power supply configuration; w is the attack strategy variable vector; z is the distribution network power flow optimization variable vector; r is the continuous variable for topology reconstruction; c is the discrete variable for topology reconstruction; s is the renewable energy output scenario; S is the set of renewable energy output scenarios; ρ s is the probability of renewable energy output scenario s; t is the optimized time period; T is the optimized time period set; It is the active power cut-off value of node j in time period t under renewable energy output scenario s.

[0074] Assuming that the line will be in a fault disconnected state during the optimization time after being attacked, attack and line state constraints are formulated, and the attack and line state constraints are used to constrain the renewable energy output scenario. The calculation formula of the attack and line state constraints is as follows:

[0075]

[0076]

[0077] in, is a binary variable representing the open and closed state of the line. When , it indicates that the circuit is in the disconnected state, otherwise it indicates that the circuit is in the closed state; u ij is a binary variable indicating whether the line (i, j) is attacked. ij =1, it means that line (i, j) is attacked; t is the optimized time period; s is the renewable energy output scenario; U is the attack uncertainty set; L is the combination of distribution network lines; B is the attack budget of the maximum number of attack lines; if the line is not attacked, the binary variable The free control takes 0 or 1. If the line is attacked, the binary variable Can only be 0.

[0078] Preferably, by taking minimizing the expected load shedding as the objective function, not only the configuration cost of distributed renewable energy sources is taken into account, but also the potential risk of load shedding is included in the optimization scope, balancing the economy and reliability of the system; by introducing attack and line status constraints, potential network attacks are taken into account. This forward-looking design enhances the resilience of the system in the face of malicious attacks and provides additional protection for the safe operation of the power grid.

[0079] S4: Build a prediction model based on the optimized renewable energy output scenario and perform prediction to obtain the power generation trend prediction results.

[0080] Preferably, constructing a prediction model and performing prediction to obtain a power generation trend prediction result includes the following steps: obtaining an original sequence based on a proximity data extraction method and performing an accumulation operation to obtain an accumulated sequence. The specific formula is as follows:

[0081] x0={x0(1),x0(2),...,x0(n)}

[0082]

[0083] Among them, x0 is the original sequence; n is the total number of original data; x1 is the cumulative sequence; x0(j) is the jth data in the original sequence; i is the number of data used in the accumulation operation.

[0084] Construct the corresponding differential formula based on the cumulative series. The specific formula is as follows:

[0085]

[0086] Among them, x1 is the cumulative series; t is the time variable; a is the development gray number; x is the variable represented by the cumulative series; u is the endogenous control gray number.

[0087] Based on the discrete formula, derivative definition, approximate formula, equal time interval method and average evaluation, the original series and the accumulated series are calculated and transformed identically. The specific formula is as follows:

[0088]

[0089] i=1,2,...,n-1

[0090] Among them, x1 is the cumulative series; a is the development gray number; b is a constant; x0 is the original series; u is the endogenous control gray number.

[0091] The formula obtained after the identity transformation is converted into a matrix expression and calculated to obtain the corresponding vector. The specific formula is as follows:

[0092]

[0093] Among them, x0 is the original sequence; n is the total number of original data; x1 is the cumulative sequence; a is the development gray number; u is the endogenous control gray number; is the parameter vector to be estimated; To develop an estimate of the gray scale; is the estimated value of the endogenous control gray number; B is the attack budget of the maximum number of attack lines; R is the correlation matrix; Y is the observation vector.

[0094] Solving equations based on predicted values ​​and vectors Construct the time response function, perform cumulative subtraction operations, and then restore the grey prediction model, that is, the prediction model. The specific formula is as follows:

[0095]

[0096] in, is the estimated value of the i-th item in the cumulative sequence x1; x1 is the cumulative sequence; is the estimated value of the endogenous control gray number; is the predicted value of the corresponding development gray number a; is the estimated value of the i-th item in the original sequence; is the estimated value of the endogenous control gray number of the i-th item; It is the estimated value of the i+1th item in the cumulative sequence x1; the predicted values ​​using the grey prediction theory all meet the standards, the prediction effect is good, and the deviation between the predicted values ​​and the actual values ​​is small, which can meet higher requirements.

[0097] It should be noted that the prediction model adopts the Gray Forecast Model. When constructing the prediction model, the gray prediction model requires a small amount of data and the information does not need to be complete. Especially when analyzing and modeling systems with less statistical data, shorter time series and incomplete information, it is also efficient and accurate. The gray prediction model mainly relies on building a mathematical model to describe the laws of the data group, and ultimately uses this to predict the specific situation of the future development of the data. It is an effective method for dealing with and analyzing small sample short-term prediction problems.

[0098] S5: Provide risk warning based on power generation trend prediction results to achieve variable grid group warning for cloud-edge collaboration.

[0099] Preferably, Figure 3 The figure shows a risk warning flow chart of a variable grid group warning method for cloud-edge collaboration. The risk warning includes the following steps: setting a risk threshold and using Sigmoid as an activation function to control whether the fault warning is triggered.

[0100] The output signal of the activation function is used as the enable signal for early warning triggering, and the adjacent value of the risk threshold is used as the dead zone.

[0101] Map the risk assessment value to the comprehensive assessment result for risk warning, realize fault warning through triggering signal, and initiate safety precautions in time to ensure safe and stable operation of the system.

[0102] Specifically, in risk warning, by using the Sigmoid function with a smooth curve as the activation function to control whether the fault warning is triggered, the input signal is mapped into an output value between 0 and 1, and whether the warning is triggered is determined according to the size of the output value; by using the adjacent value of the risk threshold as the dead zone, the setting of the dead zone can avoid the system from frequently triggering the warning when the risk value is close to the threshold, reducing the possibility of false alarms. For example, if the risk threshold is 30, the dead zone can be set to the range of ±5 of the threshold. When the risk assessment value is between 25 and 35, the system will not trigger the warning. When the risk assessment value exceeds this range, the warning will be triggered.

[0103] Specifically, fault warning is achieved by using a trigger signal. When the output signal of the activation function reaches a certain level, the warning mechanism is triggered. This trigger signal can be an electrical signal, a software signal or a signal in other forms. Safety precautions are initiated by utilizing the trigger signal. Once the warning is triggered, the system needs to take corresponding measures immediately, such as shutting down some equipment, adjusting operating parameters or issuing alarms, thereby reducing risks and preventing faults from occurring.

[0104] Furthermore, risk assessment values ​​are calculated by utilizing sensor data, historical data, and expert knowledge.

[0105] Specifically, the risk assessment value can be assessed by the three-point estimation method, or by other methods, which are mainly divided into simulation method and analytical method, wherein the simulation method includes Monte Carlo method, LHS sampling method, non-sequential simulation method and sequential simulation method, and the analytical method includes event tree method, fault tree method and semi-invariant method, etc., and the present invention mainly considers the node voltage over-limit risk value and the line power overload risk value, wherein the node voltage over-limit risk value and the line power overload risk value can be obtained by expert knowledge or historical data statistics, that is, the risk over-limit probability and over-limit value of the node voltage over-limit risk value and the line power overload risk value are statistically obtained, and can also be calculated using a formula, and the risk assessment value can be obtained by normalizing the node voltage over-limit risk value and the line power overload risk value, or a single risk value can be used as an indicator.

[0106] It should be noted that the node voltage over-limit risk value refers to the deviation between the voltage of each node in the distribution network and the specified range. When the voltage deviation exceeds the specified range of ±5%, it means that the node voltage is in an over-limit state. The specific formula of the node voltage over-limit risk value is as follows:

[0107]

[0108] in, is the voltage over-limit risk value of node k; is the lower limit of voltage allowed; is the upper limit of voltage allowed; f(Vk ) is the voltage probability density function of node k; Sev vl (V k ) is the severity function of the voltage exceeding the lower limit at node k; Sev vh (V k ) is a function of the severity of the voltage exceeding the upper limit at node k.

[0109] It should be noted that the line power overload risk value refers to whether the active power carried by the line exceeds the rated maximum capacity limit under different system operating conditions of the distribution network, which may cause overload risks. The specific formula of the line power overload risk value is as follows:

[0110]

[0111] in, is the power overload risk value of line l; is the upper limit of line power allowed; f(P l ) is the probability density function of line power; Sev(P l ) is the function of the severity of line power exceeding the limit.

[0112] Preferably, by adopting the Sigmoid function as the activation function for early warning triggering, the nonlinear changes of the actual risk situation can be better simulated, and the sensitivity of the early warning can be flexibly controlled; by setting the adjacent value of the risk threshold as the dead zone, the frequent response of the system when the risk value fluctuates near the critical point is avoided; by adopting a comprehensive evaluation method that considers the node voltage over-limit risk and the line power overload risk, and by introducing the severity function, the degree of impact after the risk occurs is considered, providing a more comprehensive risk assessment.

[0113] In summary, the present invention configures renewable energy power sources in the planning stage for the most serious fault scenario under the fault uncertainty set, thereby improving the resilience of the distribution network; by considering the impact of renewable energy output uncertainty, a series of renewable energy output scenarios are constructed, and the optimal configuration is performed with the goal of minimizing the expected load shedding in each scenario, thereby improving the robustness of the prediction.

[0114] Embodiment 2 is an embodiment of the present invention, which provides a variable grid group early warning system for cloud-edge collaboration, including: a characteristic function module, which is used to obtain data on corresponding climate factors under complex terrain and construct characteristic functions of different climate factors on wind field airflow fields; a scenario construction module, which is used to construct renewable energy output scenarios and perform scenario generation and scenario reduction according to wind speed uncertainty; a scenario optimization module, which is used to optimize the configuration of renewable energy output scenarios and formulate attack and line status constraints; a trend prediction module, which is used to construct a prediction model based on the optimized renewable energy output scenario and perform prediction to obtain power generation trend prediction results; a risk warning module, which is used to make judgments based on power generation trend prediction results and risk thresholds to achieve variable grid group early warning for cloud-edge collaboration.

[0115] Embodiment 3 is an embodiment of the present invention, which is different from the previous embodiment in that:

[0116] like Figure 4 As shown, if the function is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art or the part of the technical solution, can be embodied in the form of a software product, which is stored in a storage medium and includes several instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk and other media that can store program codes.

[0117] The logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as an ordered list of executable instructions for implementing logical functions, and can be embodied in any computer-readable medium for use by an instruction execution system, device or apparatus (such as a computer-based system, a system including a processor, or other system that can fetch instructions from an instruction execution system, device or apparatus and execute instructions), or in conjunction with such instruction execution systems, devices or apparatuses. For the purposes of this specification, "computer-readable medium" can be any device that can contain, store, communicate, propagate or transmit a program for use by an instruction execution system, device or apparatus, or in conjunction with such instruction execution systems, devices or apparatuses.

[0118] More specific examples of computer-readable media (a non-exhaustive list) include the following: an electrical connection with one or more wires (electronic device), a portable computer disk case (magnetic device), a random access memory (RAM), a read-only memory (ROM), an erasable and programmable read-only memory (EPROM or flash memory), an optical fiber device, and a portable compact disk read-only memory (CDROM). In addition, the computer-readable medium may even be a paper or other suitable medium on which the program is printed, since the program may be obtained electronically, for example, by optically scanning the paper or other medium, followed by editing, deciphering or, if necessary, processing in another suitable manner, and then stored in a computer memory.

[0119] It should be understood that the various parts of the present invention can be implemented by hardware, software, firmware or a combination thereof. In the above-mentioned embodiments, a plurality of steps or methods can be implemented by software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented by hardware, as in another embodiment, it can be implemented by any one of the following technologies known in the art or their combination: a discrete logic circuit having a logic gate circuit for implementing a logic function for a data signal, a dedicated integrated circuit having a suitable combination of logic gate circuits, a programmable gate array (PGA), a field programmable gate array (FPGA), etc.

[0120] Example 4 is an embodiment of the present invention, which provides a variable grid group early warning method for cloud-edge collaboration. In order to verify the beneficial effects of the present invention, scientific demonstration is carried out through simulation experiments.

[0121] This example uses the CPS62 node example to study the impact of double-end uncertainty of wind turbines, photovoltaics and loads on the operation safety of various areas of the distribution network, including three areas: industrial area, commercial area and residential area, with a total of 62 nodes, 59 branches, 6 tie lines, including three transformers S1, S2 and S3. Figure 5 shown.

[0122] Among them, the daily load curves of residential areas, industrial areas and commercial areas have obvious time characteristics, such as Figure 6 The following are typical daily load curves for various regions. It can be seen from the figure that the load in residential areas is at two peaks in the two time periods of 10-12 o'clock and 18-22 o'clock, and reaches the highest peak at 19 o'clock. Due to the influence of peak-valley electricity prices, the load in industrial areas is at its peak from 22:00 to 6:00, with a large electricity load, and is in the trough period from 10:00 to 18:00. The load in commercial areas begins to gradually increase from 7:00 to 12:00, is at its peak from 12:00 to 19:00, and begins to decline after 19:00. Due to the different natures of the regions, and the fact that the daily load curves of the three regions have obvious time characteristics, the operation risks of the three regions will show obvious temporal and spatial differences.

[0123] like Figure 7 to Figure 9 The figure shows the comprehensive risk value of voltage over-limit at each node in each area, the comprehensive risk value of power overload on each line in each area, and the comprehensive risk value of each area. It can be seen from the figure that the operation risk in the residential area is relatively high during the two periods of 12-14 o'clock and 18-20 o'clock; from 12-14 o'clock, due to the increase in photovoltaic output and the relatively small load, the power is delivered, and the risk of voltage exceeding the upper limit increases; from 18-20 o'clock, the load in the residential area is at its peak. Since there is no sunlight at this time, the photovoltaic output decreases, resulting in the voltage exceeding the lower limit and the risk of line flow overload increases; the operation risk in the commercial area is relatively high between 9-10 o'clock and 18-19 o'clock. From 9-10 o'clock, the load in the commercial area and the photovoltaic output in the commercial area are in the process of increasing at the same time. Due to the excessive increase in photovoltaic output, the risk of voltage exceeding the lower limit and line flow overload increases. The fast operation leads to an increase in the risk of voltage exceeding the upper limit. From 18:00 to 19:00, the load is at its peak and the photovoltaic output decreases, resulting in insufficient power support, and the risk of voltage exceeding the lower limit and power flow overload increases. The operation risk of the industrial zone is relatively large from 1:00 to 6:00 and from 11:00 to 15:00. From 1:00 to 6:00, the load in the industrial zone is at its peak. At the same time, the installed capacity of wind turbines is small, and the power support is insufficient. The risk of voltage exceeding the lower limit and power flow overload increases. From 11:00 to 15:00, the load in the industrial zone is at a low period, and the photovoltaic output is at its peak, resulting in an increase in the risk of voltage exceeding the upper limit. This shows that the present invention effectively improves the robustness of the prediction by considering the influence of the uncertainty of renewable energy output and optimizing the configuration with the minimum expected load shedding in each scenario as the goal.

[0124] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present invention, which should all be included in the scope of the claims of the present invention.

Claims

1. A variable grid group early warning method for cloud-edge collaboration, characterized by: include: Obtain data on climate factors and construct characteristic functions of different climate factors on wind field, air flow and flow field; Construct renewable energy output scenarios and perform scenario generation and scenario reduction; Optimizing the renewable energy output scenario; Build a prediction model based on the optimized renewable energy output scenario and make predictions to obtain the power generation trend prediction results; Risk warning is performed based on the power generation trend prediction results to achieve variable grid group warning for cloud-edge collaboration.

2. The variable grid group early warning method for cloud-edge collaboration according to claim 1, characterized in that: The data of climate factors are composed of data sets of climate factors corresponding to complex terrain and landforms; The construction of characteristic functions of different climate factors on wind field airflow flow field comprises the following steps: Obtain data on climate factors; According to the data of the climate factors, a mapping relationship between the climate factors and the wind field, air flow field and flow field is constructed; Based on the mapping relationship between the climate factors and the wind farm airflow field, characteristic functions of different climate factors with respect to the wind farm airflow field are constructed.

3. The variable grid group early warning method for cloud-edge collaboration according to claim 2 is characterized in that: The renewable energy output scenario includes: Aiming at the uncertainty of wind speed, Weibull distribution is used to describe the distribution of wind field airflow. Describe the relationship between wind speed and output power; The scene generation refers to performing stratified sampling of wind speed using the Latin hypercube sampling method to generate a wind speed sequence scene; The scene reduction refers to determining a subset of the initial scene and reallocating probabilities for the subset to obtain a subset after allocation, so that the probability distance between the probability distribution Q of the subset after allocation and the probability distribution P of the subset of the initial scene is the shortest.

4. The variable grid group early warning method for cloud-edge collaboration according to claim 3 is characterized in that: Optimizing the renewable energy output scenario includes: Taking the minimum load shedding expectation in each scenario as the goal, optimizing the configuration of the renewable energy output scenario; Assuming that the line will be in a fault disconnected state for an optimized time after being attacked, attack and line state constraints are formulated, and the renewable energy output scenario is constrained by using the attack and line state constraints.

5. The variable grid group early warning method for cloud-edge collaboration according to claim 4 is characterized in that: The method of constructing a prediction model and performing prediction to obtain a power generation trend prediction result includes the following steps: Based on the adjacent data extraction method, the original sequence is obtained and an accumulation operation is performed to obtain an accumulated sequence; Constructing a corresponding differential expression based on the accumulated sequence; Based on the discrete formula, the derivative definition, the approximate formula, the equal time interval method and the average evaluation, the original number series and the accumulated number series are calculated and identically transformed; The formula obtained after the identity transformation is converted into a matrix expression and calculated to obtain the corresponding vector; Solving equations based on predicted values ​​and vectors Construct the time response function, perform cumulative subtraction operation, and then restore the grey prediction model, that is, the prediction model; The risk warning comprises the following steps: Set the risk threshold and use Sigmoid as the activation function to control whether the fault warning is triggered; The output signal of the activation function is used as the enable signal for early warning triggering, and the value near the risk threshold is used as the dead zone; Map the risk assessment value to the comprehensive assessment result for risk warning, realize fault warning through triggering signal, and initiate safety precautions in time.

6. The variable grid group early warning method for cloud-edge collaboration according to claim 5, characterized in that: The calculation formula of the wind field airflow distribution is as follows: Among them, f w (v) is the probability density function representing wind speed; k is the shape parameter; c is the scale parameter; v is the wind speed; The calculation formula for the relationship between wind speed and output power is as follows: Among them, P w is the output power; v is the wind speed; v ci is the cut-in wind speed; v rate is the rated wind speed; v co is the cut-out wind speed; P rate is the rated capacity of the wind turbine; The calculation formula of the probability distance between the probability distribution Q of the assigned subset and the probability distribution P of the subset of the initial scene is as follows: Among them, D k is the probability distance between the probability distribution Q of the assigned subset and the probability distribution P of the subset of the initial scene; P is the probability distribution of the subset of the initial scene; Q is the probability distribution of the assigned subset; S is the number of scenes in the probability distribution P of the subset of the initial scene; is the number of scenarios of the probability distribution Q of the allocated subset; η ij is the decision variable in the model; For the scene i and scenes The distance between i is a scene in the probability distribution P of a subset of the initial scenes; p i For the scene i Probability of occurrence; is the scenario in the probability distribution Q of the assigned subset; q j For the scene The probability of occurrence; k is the shape parameter.

7. The variable grid group early warning method for cloud-edge collaboration according to claim 6 is characterized in that: The calculation formula of the objective function for optimizing the configuration of the renewable energy output scenario is as follows: Among them, j is a node; V is a node set; h is the cost of configuring distributed renewable energy sources at node j; j is the binary decision variable for whether to configure power supply at node j; h is the decision variable for distributed renewable energy power supply configuration; w is the attack strategy variable vector; z is the distribution network power flow optimization variable vector; r is the continuous variable for topology reconstruction; c is the discrete variable for topology reconstruction; s is the renewable energy output scenario; S is the set of renewable energy output scenarios; ρ s is the probability of renewable energy output scenario s; t is the optimized time period; T is the optimized time period set; is the active power cut-off value of node j in time period t under renewable energy output scenario s; The calculation formula of the attack and line state constraint is as follows: in, is a binary variable representing the open and closed state of the line; u ij is a binary variable indicating whether the line (i, j) is attacked; t is the optimized time period; s is the renewable energy output scenario; U is the attack uncertainty set; L is the combination of distribution network lines; B is the attack budget for the maximum number of attack lines.

8. A variable grid group early warning system for cloud-edge collaboration, based on a variable grid group early warning method for cloud-edge collaboration according to any one of claims 1 to 7, characterized in that: include, The characteristic function module is used to obtain the data of the corresponding climate factors under complex terrain and to construct the characteristic functions of different climate factors on the wind field and air flow field; Scenario building module, used to build renewable energy output scenarios and perform scenario generation and scenario reduction based on wind speed uncertainty; Scenario optimization module, used to optimize the configuration of renewable energy output scenarios and formulate attack and line status constraints; The trend prediction module is used to build a prediction model based on the optimized renewable energy output scenario and perform predictions to obtain power generation trend prediction results; The risk warning module is used to make judgments based on power generation trend prediction results and risk thresholds to achieve variable grid group warnings for cloud-edge collaboration.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of a variable grid group early warning method for cloud-edge collaboration as described in any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of a variable grid group early warning method for cloud-edge collaboration as described in any one of claims 1 to 7 are implemented.

Images