Power grid time sequence operation risk prediction method and system
The method and system integrate power system dispatching, risk assessment, and artificial intelligence to predict power grid operation risks, addressing inaccuracies in existing technologies by using an AC power grid optimization scheduling model and deep neural networks for comprehensive risk evaluation and prediction.
Patent Information
- Application Number
- CN202410904922.9
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-07-08
- Publication Date
- 2024-12-06
AI Technical Summary
Existing power grid risk assessment technologies fail to accurately quantify and predict risks such as load shedding, voltage over-limit, and branch current over-limit due to the randomness and intermittency of renewable energy, lacking comprehensive evaluation methods and integration of power system dispatching, risk assessment, and artificial intelligence.
A method and system using an AC power grid optimization scheduling model with second-order cone relaxation, combined with a risk assessment system and deep neural network, to evaluate and predict power grid operation risks by considering frequency margin, capacity reserve, line overload, and voltage deviation, and normalizing indicators with a Gaussian mixture model.
Accurately predicts future power grid operation risks by integrating multiple safety indicators and data-driven methods, providing a reliable reference for forward-looking dispatching with improved evaluation accuracy.
Smart Images

Figure CN119090253A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of power grid operation risk assessment and prediction, and in particular to a power grid sequential operation risk prediction method and system. Background Art
[0002] With the large-scale access of distributed renewable energy, the operational safety issues of power systems have increased significantly in recent years. The randomness and intermittency of renewable energy make the state changes of power systems more complicated, which may cause problems such as load shedding, voltage over-limit, and branch current over-limit. Risk assessment can quantify the risks of power systems and is a necessary link and reliable guarantee for power system safety analysis. Especially in the process of forward-looking dispatching of power grids, it is necessary to grasp the future short-term operational risk status. In this context, it is necessary to carry out predictions for the time-series operation risks of power grids, and provide risk analysis and early warning means for control personnel by perceiving the operation trends and operation risks of power grids in advance, so as to realize multi-time scale power grid pre-dispatch.
[0003] Based on this, the research on grid timing operation risk prediction based on deep neural networks needs to fully consider the following two aspects: First, for the construction of the current risk assessment system, it is necessary to comprehensively consider multiple factors that affect the stability of grid operation and adopt appropriate assignment methods to achieve risk quantification; Second, compared with traditional risk assessment and prediction technologies, how to combine power system dispatching theory, risk assessment theory and artificial intelligence theory to achieve accurate prediction of grid operation risks. Summary of the invention
[0004] In view of the above existing problems, the present invention is proposed.
[0005] Therefore, the present invention provides a method and system for predicting the risks of power grid timing operation to solve the problems that the risks of load shedding, voltage exceeding the limit, and branch current exceeding the limit caused by system state quantity switching cannot be quantitatively evaluated, and the evaluation accuracy is insufficient.
[0006] In order to solve the above technical problems, the present invention provides the following technical solutions:
[0007] In a first aspect, the present invention provides a method for predicting the risk of power grid timing operation, comprising: obtaining power grid operation parameters, constructing an AC power grid optimization scheduling model based on second-order cone relaxation according to the power grid operation parameters, wherein the AC power grid optimization scheduling model takes the lowest overall operating cost as the objective function and takes into account relevant constraints of power grid operation; based on the AC power grid optimization scheduling model, considering four indicators including frequency margin, capacity reserve, line overload and voltage deviation, constructing a risk assessment system, and using the risk assessment system to assign values to and evaluate the power grid operation risks; constructing a deep neural network training model through the risk assessment data of the assigned assessment, taking the risk assessment data of four consecutive time sections as the input of the deep neural network training model, outputting the risk assessment data of a future time section, and realizing the prediction of the risk situation of the future time section by the risk assessment data of the current time section.
[0008] As a preferred solution of the method for predicting the time sequence operation risk of the power grid described in the present invention, the construction of the AC power grid optimization dispatching model based on second-order cone relaxation includes:
[0009] According to the obtained grid operation parameters, the second-order cone relaxation of the nonlinear terms of the AC grid is adopted. The overall operation cost is minimized as the objective function. The reconstruction constraints of the nonlinear terms of the grid operation, the node power balance constraints, the branch flow constraints, the generator set operation status constraints, and the generator set output constraints are taken into account to construct an AC grid optimization scheduling model.
[0010] As a preferred solution of the method for predicting the risk of power grid timing operation described in the present invention, the objective function of the AC power grid optimization dispatching model is expressed as:
[0011]
[0012] Among them, h v,t (z v,t ,u v,t ,v v,t ) represents the fixed cost of generator set v at time t, c v represents the unit power generation cost of generator v at time t, z v,t represents an integer variable, which is the operating status of the generator set v at time t, u v,t represents an integer variable, which is the starting state of the generator set v at time t, v v,t represents an integer variable, which is the shutdown state of the generator set v at time t, Represents the active output of generator set v at time t.
[0013] As a preferred solution of the method for predicting the risk of sequential operation of a power grid according to the present invention, the risk assessment system is constructed as follows:
[0014] Considering the four indicators of frequency margin, capacity reserve, line overload and voltage deviation, they are characterized by maximum disturbance power, reserve power, line overload rate and voltage deviation rate respectively;
[0015] The frequency margin is calculated as:
[0016]
[0017]
[0018] Where ΔP max represents the maximum disturbance power, Δf max represents the maximum disturbance frequency, D represents the load damping constant, R represents the speed regulator adjustment constant, ω n , α, ω r , t nadir All represent intermediate variables and have no clear physical meaning. H represents the system inertia constant, T R represents the time constant, F H represents the generator turbine parameters;
[0019] The capacity reserve is calculated as:
[0020]
[0021] in, and They represent the upper reserve power and lower reserve power provided by the generator set v at time t, represents the total reserve power of the power grid at time t, Indicates the maximum climbing power of the generator set v, and They respectively represent the lower and upper limits of the active output of the generator set v.
[0022] As a preferred solution of the method for predicting the risk of power grid sequential operation according to the present invention, the risk assessment system construction further includes:
[0023] The calculation of the line overload is:
[0024] β=n br / N br
[0025] Where β represents the line overload rate, n br Indicates the number of lines whose transmission power exceeds 80% of the transmission capacity, N br Indicates the total number of lines;
[0026] The voltage deviation is calculated as:
[0027] γ=n node / N node
[0028] Among them, γ represents the voltage deviation rate, n node Indicates the number of nodes whose voltage fluctuation exceeds 0.05V, N node Indicates the total number of nodes.
[0029] As a preferred solution of the method for predicting the time sequence operation risk of the power grid according to the present invention, the evaluation of the power grid operation risk includes:
[0030] The values of each indicator under the risk assessment system under given scheduling data are normalized and then modeled using a Gaussian mixture model, which is expressed as follows:
[0031]
[0032]
[0033] Among them, F(y) represents the Gaussian mixture model, k represents a single Gaussian component, K represents the number of Gaussian components, α k , μ k and Respectively represent the weight coefficient, expectation and variance of Gaussian component k, represents the probability density function of Gaussian component k, and y represents the normalized indicator data;
[0034] Specify a confidence level and give a corresponding score table, and obtain the confidence interval based on the specified confidence level:
[0035]
[0036] Among them, ε represents the confidence level, δ up and δ dn Represent the upper and lower bounds of the confidence interval respectively;
[0037] Each indicator under the risk assessment system is assigned a value based on the interval in which it is located:
[0038]
[0039] Among them, y(t) represents the indicator data at time t, E(t) = {E P (t),E R (t),E L (t),E V (t)} represents the specific values of the four indicators of frequency margin, capacity reserve, line overload and voltage deviation at time t. Represents an assignment function.
[0040] As a preferred solution of the method for predicting the risk of power grid sequential operation according to the present invention, the risk assessment data of the current time section is used to predict the risk of the future time section, which includes:
[0041] A risk assessment system training data is generated for the risk assessment data of each indicator after the assigned evaluation, and a deep neural network training model is constructed based on the risk assessment system training data. After the training is completed, the risk assessment data of four consecutive time sections are used as the input of the deep neural network training model, and the risk assessment data of a future time section is output to predict the risk situation of the future section based on the risk assessment data of the current section.
[0042] In a second aspect, the present invention provides a power grid sequential operation risk prediction system, comprising:
[0043] A parameter acquisition unit, used to acquire power grid operation parameters;
[0044] A model building module, used to build an AC power grid optimization dispatching model based on second-order cone relaxation according to the power grid operation parameters, wherein the AC power grid optimization dispatching model takes the lowest overall operation cost as the objective function and takes into account the relevant constraints of the power grid operation;
[0045] A value evaluation module is used to construct a risk evaluation system based on the AC power grid optimization dispatching model, taking into account four indicators: frequency margin, capacity reserve, line overload and voltage deviation, and use the risk evaluation system to perform value evaluation on power grid operation risks;
[0046] The risk prediction module is used to construct a deep neural network training model through the risk assessment data of the assigned evaluation, use the risk assessment data of four consecutive time sections as the input of the deep neural network training model, output the risk assessment data of a future time section, and realize the prediction of the risk situation of the future time section by the risk assessment data of the current time section.
[0047] In a third aspect, the present invention provides a computing device, comprising:
[0048] Memory and processor;
[0049] The memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions. When the computer-executable instructions are executed by the processor, the steps of the method are implemented.
[0050] In a fourth aspect, the present invention provides a computer-readable storage medium storing computer-executable instructions, which implement the steps of the method when executed by a processor.
[0051] Compared with the prior art, the beneficial effects of the present invention are as follows: the present invention comprehensively considers multiple factors that affect the stability of power grid operation, combines multiple safety indicators with data-driven assignment methods to quantify power grid operation risks, and predicts future operation risks based on the current risk status; compared with traditional risk assessment and prediction technologies, the present invention combines power system dispatching theory, risk assessment theory and artificial intelligence theory to achieve accurate prediction and safety assessment of power grid operation risks, and provide an operation reference for forward-looking power grid dispatching. BRIEF DESCRIPTION OF THE DRAWINGS
[0052] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other accompanying drawings can be obtained based on these accompanying drawings without paying creative work.
[0053] Figure 1 This is an overall flow chart of a method for predicting the risk of power grid sequential operation according to an embodiment of the present invention;
[0054] Figure 2 A schematic diagram of a deep neural network training model for a method for predicting the risk of power grid timing operation according to an embodiment of the present invention;
[0055] Figure 3 A frequency distribution histogram and a schematic diagram of a fitting curve of various indicators of a method for predicting the risk of sequential operation of a power grid according to an embodiment of the present invention;
[0056] Figure 4 This is a comparison chart between actual values and predicted values of a risk assessment system of a method for predicting risks of sequential operation of a power grid according to an embodiment of the present invention. DETAILED DESCRIPTION
[0057] In order to make the above-mentioned purposes, features and advantages of the present invention more obvious and easy to understand, the specific implementation methods of the present invention are described in detail below in conjunction with the drawings of the specification. Obviously, the described embodiments are part of the embodiments of the present invention, but not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary persons in the art without creative work should fall within the scope of protection of the present invention.
[0058] Example 1
[0059] Reference Figure 1-2 According to an embodiment of the present invention, a method for predicting the risk of power grid sequential operation is provided, which specifically comprises the following steps:
[0060] S100: Obtaining grid operation parameters;
[0061] Optionally, the acquired power grid operation parameters include network topology, node admittance matrix, node voltage, branch flow, generator set, load and new energy unit parameter information.
[0062] S200: constructing an AC power grid optimization dispatching model based on second-order cone relaxation according to the power grid operation parameters, wherein the AC power grid optimization dispatching model takes the minimum overall operation cost as the objective function and takes into account the relevant constraints of the power grid operation;
[0063] In the embodiment of the present application, according to the acquired grid operation parameters, the nonlinear terms of the AC grid are relaxed by a second-order cone, the overall operation cost is minimized as the objective function, and the nonlinear term reconstruction constraints, node power balance constraints, branch power flow constraints, generator set operation status constraints, and generator set output constraints of the grid operation are taken into account to construct an AC grid optimization scheduling model;
[0064] Specifically, the calculation of the nonlinear reconstruction constraint is:
[0065] m ii,t =|V i,t | 2
[0066] m ij,t =|V i,t ||V j,t |cos(θ i,t -θ i,t )
[0067] n ij,t =-|V i,t ||V j,t |sin(θ i,t -θ j,t )
[0068]
[0069] (m ij,t ) 2 +(n ij,t ) 2 ≤m ii,t m jj,t
[0070] n ij,t =tan(θ j,t -θ i,t )m ij,t
[0071] Where, subscript t represents the dispatch time, subscripts i and j represent power nodes, V i,t and V j,t They represent the voltage of node i at time t, θ i,t and θi,t They represent the voltage phase angle of node i at time t, m ii,t ,m ij,t ,n ij,t is the reconstructed variable, used to characterize the nonlinear term of voltage, v i and They represent the lower and upper limits of the voltage amplitude at node i, respectively, v j and They represent the lower and upper limits of the voltage amplitude at node j, θ ij and They represent the lower limit and upper limit of the phase angle difference of branch ij at time t respectively;
[0072] Specifically, the calculation of the node power balance constraint is:
[0073]
[0074] Where, the subscript v represents the generator set, σ v (i) represents the set of generators connected to node i, σ(i) represents the set of nodes connected to node i, and They represent the active and reactive output of the generator set v at time t, respectively. and They represent the active power demand and reactive power demand of node i at time t, and They represent the active power and reactive power flowing out of line ij at time t, and They represent the active power and reactive power flowing into line ij at time t, respectively, and g ii and b ii denote the self-conductance and self-susceptance of node i respectively;
[0075] Specifically, the calculation of branch power flow constraint is:
[0076]
[0077] Among them, G ij and B ij denote the real and imaginary parts of the node admittance matrix respectively;
[0078] Specifically, the operating status constraints of the generator set are:
[0079] z v,t-1 -z v,t =u v,t-1 -v v,t
[0080] u v,t -v v,t ≤0
[0081] v v,t +z v,t ≤1
[0082]
[0083]
[0084] Among them, z v,t is an integer variable, indicating the operating status of the generator set v at time t, u v,t and v v,t are integer variables, representing the start and shutdown status of the generator set v at time t, and They represent the minimum start-up time and minimum shutdown time of the generator set v respectively;
[0085] Specifically, the calculation of the generator set output constraint is:
[0086]
[0087] in, and They represent the lower and upper limits of the active output of the generator set v, respectively. and They represent the lower and upper limits of reactive power output of generator set v, respectively. Indicates the maximum climbing power of the generator set v;
[0088] Specifically, the objective function of the AC power grid optimization dispatching model is expressed as:
[0089]
[0090] Among them, h v,t (z v,t ,u v,t ,v v,t ) represents the fixed cost of generator set v at time t, c v represents the unit power generation cost of generator v at time t, z v,t represents an integer variable, which is the operating status of the generator set v at time t, u v,t represents an integer variable, which is the starting state of the generator set v at time t, v v,t represents an integer variable, which is the shutdown state of the generator set v at time t, Represents the active output of generator set v at time t.
[0091] S300: Based on the AC power grid optimization dispatching model, considering the four indicators of frequency margin, capacity reserve, line overload and voltage deviation, a risk assessment system is constructed, and the risk assessment system is used to assign values to the power grid operation risks;
[0092] In the embodiment of the present application, the four indicators of frequency margin, capacity reserve, line overload and voltage deviation are considered as the evaluation indicators of the present invention, and then the four indicators are characterized and quantified by the maximum disturbance power, reserve power, line overload rate and voltage deviation rate respectively according to the power grid dispatching data;
[0093] Specifically, the frequency margin is calculated as:
[0094]
[0095]
[0096] Where ΔP max represents the maximum disturbance power, Δf max represents the maximum disturbance frequency, D represents the load damping constant, R represents the speed regulator adjustment constant, ω n , α, ω r , t nadir All represent intermediate variables and have no clear physical meaning. H represents the system inertia constant, T R represents the time constant, F H represents the generator turbine parameters;
[0097] Specifically, the calculation of capacity reserve is:
[0098]
[0099] in, and They represent the upper reserve power and lower reserve power provided by the generator set v at time t, represents the total reserve power of the power grid at time t, Indicates the maximum climbing power of the generator set v, and They represent the lower and upper limits of the active output of the generator set v respectively;
[0100] Specifically, the calculation of line overload is:
[0101] β=n br / N br
[0102] Where β represents the line overload rate, n br Indicates the number of lines whose transmission power exceeds 80% of the transmission capacity, N br Indicates the total number of lines;
[0103] Specifically, the voltage deviation is calculated as:
[0104] γ=n node / N node
[0105] Among them, γ represents the voltage deviation rate, n node Indicates the number of nodes whose voltage fluctuation exceeds 0.05V, N node Indicates the total number of nodes;
[0106] Furthermore, the values of each indicator under the risk assessment system under given scheduling data are normalized, and the formula is expressed as follows:
[0107]
[0108] in, Indicates the specific value of each indicator under given scheduling data, x min and x max They represent the minimum and maximum values of the corresponding indicators respectively, and y represents the normalized indicator data;
[0109] Then the Gaussian mixture model is used for modeling, and the formula is expressed as:
[0110]
[0111] Among them, F(y) represents the Gaussian mixture model, k represents a single Gaussian component, K represents the number of Gaussian components, α k , μ k and Respectively represent the weight coefficient, expectation and variance of Gaussian component k, represents the probability density function of Gaussian component k, and y represents the normalized indicator data;
[0112] Furthermore, specify the confidence level and give the corresponding score table, and obtain the confidence interval according to the specified confidence level:
[0113]
[0114] Among them, ε represents the confidence level, δ up and δ dn Represent the upper and lower bounds of the confidence interval respectively;
[0115] Furthermore, each indicator under the risk assessment system is assigned a value based on the interval in which it is located:
[0116]
[0117] Among them, y(t) represents the indicator data at time t, E(t) = {E P(t),E R (t),E L (t),E V (t)} represents the specific values of the four indicators of frequency margin, capacity reserve, line overload and voltage deviation at time t. Represents an assignment function.
[0118] S400: Construct a deep neural network training model through the risk assessment data of the assigned evaluation, use the risk assessment data of four consecutive time sections as the input of the deep neural network training model, output the risk assessment data of a future time section, and realize the prediction of the risk situation of the future time section by the risk assessment data of the current time section.
[0119] The risk assessment data of each indicator after the assigned evaluation are used to generate risk assessment system training data, and a deep neural network training model is constructed based on the risk assessment system training data. After the training is completed, the risk assessment data of four consecutive time sections are used as the input of the deep neural network training model, and the risk assessment data of a future time section is output to predict the risk situation of the future section based on the risk assessment data of the current section.
[0120] It can be seen from the above embodiments that the present invention comprehensively considers multiple factors that affect the stability of power grid operation, combines multiple safety indicators and data-driven assignment methods to quantify power grid operation risks, and predicts future operation risks based on the current risk status; compared with traditional risk assessment and prediction technologies, the present invention combines power system scheduling theory, risk assessment theory and artificial intelligence theory to achieve accurate prediction and safety assessment of power grid operation risks, and provide an operation reference for forward-looking power grid scheduling.
[0121] The above is a schematic scheme of a method for predicting the risk of grid timing operation in this embodiment. It should be noted that the technical scheme of the system for predicting the risk of grid timing operation is the same as the technical scheme of the method for predicting the risk of grid timing operation mentioned above. For details not described in detail in the technical scheme of the system for predicting the risk of grid timing operation in this embodiment, please refer to the description of the technical scheme of the method for predicting the risk of grid timing operation mentioned above.
[0122] This embodiment provides a power grid timing operation risk prediction system, including:
[0123] A parameter acquisition unit is used to acquire power grid operation parameters, wherein the power grid operation parameters include network topology, node admittance matrix, node voltage, branch flow, generator set, load and new energy unit parameter information;
[0124] A model building module is used to build an AC power grid optimization dispatching model based on second-order cone relaxation according to power grid operation parameters, wherein the AC power grid optimization dispatching model takes the minimum overall operation cost as the objective function and takes into account the relevant constraints of power grid operation;
[0125] The value assessment module is used to construct a risk assessment system based on the AC power grid optimization dispatch model, taking into account the four indicators of frequency margin, capacity reserve, line overload and voltage deviation, and use the risk assessment system to perform value assessment on the power grid operation risk;
[0126] The risk prediction module is used to build a deep neural network training model through the risk assessment data of the assigned evaluation, take the risk assessment data of four consecutive time sections as the input of the deep neural network training model, output the risk assessment data of a future time section, and realize the prediction of the risk situation of the future time section by the risk assessment data of the current time section.
[0127] This embodiment also provides a computing device, including:
[0128] Memory and processor;
[0129] The memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions to implement the method proposed in the above embodiment.
[0130] This embodiment also provides a storage medium on which a computer program is stored. When the program is executed by a processor, the method proposed in the above embodiment is implemented.
[0131] The storage medium proposed in this embodiment and the method proposed in the above embodiment belong to the same inventive concept. The technical details not fully described in this embodiment can be referred to the above embodiment, and this embodiment has the same beneficial effects as the above embodiment.
[0132] Through the above description of the implementation methods, the technicians in the relevant field can clearly understand that the present invention can be implemented by means of software and necessary general hardware, and of course can also be implemented by hardware, but in many cases the former is a better implementation method. Based on such an understanding, the technical solution of the present invention is essentially or the part that contributes to the prior art can be embodied in the form of a software product, and the computer software product can be stored in a computer-readable storage medium, such as a computer floppy disk, read-only memory (ReadOnly, Memory, ROM), random access memory (RandomAccess Memory, RAM), flash memory (FLASH), hard disk or optical disk, etc., including a number of instructions for a computer device (which can be a personal computer, server, or network device, etc.) to perform the methods of various embodiments of the present invention.
[0133] Example 2
[0134] Reference Figure 3 and Figure 4 Based on the previous embodiment, this embodiment provides an application example of a method and system for predicting the risk of power grid timing operation, in order to verify and illustrate the technical effects adopted in this method.
[0135] This example is based on the operation data of a certain area of Jiangsu Province in China to carry out operation scheduling research and risk prediction. The total installed capacity of the power grid is 79.86GW, of which the installed capacity of new energy generators is 18.02GW. The scheduling cycle is 168h (1 week) and the scheduling interval is 15min. Figure 2 The figure shows a schematic diagram of the deep neural network training model of this embodiment, where the model input is the risk score at four consecutive time sections, and the output is the risk score at a future time section.
[0136] like Figure 3 The frequency distribution histogram and fitting curve of each indicator are shown in the figure. Figure 3 It can be seen that the probability distribution of each indicator almost shows multi-peak and skewed irregular characteristics, and the Gaussian mixture model is composed of multiple Gaussian distributions superimposed, which can better respond to this situation. Among them, the voltage deviation rate and line overload rate are inferior indicators, and the larger the indicator value, the higher the operating risk; the spare capacity and frequency margin are superior indicators, and the larger the indicator value, the lower the operating risk.
[0137] like Figure 4 The figure shows the comparison between the actual value and the predicted value of the risk assessment system. It can be seen from the figure that from the perspective of the overall assessment system, the prediction accuracy is relatively high, and only a few indicators have obvious differences. Specifically, in comparison one, only the capacity reserve has a large difference, and the predicted values and actual values of the other three indicators are almost the same; in comparison three, only the line overload has a significant difference.
[0138] Therefore, it can be seen from the above that the present invention pays attention to the dimensional and numerical differences of various indicators, and proposes a data-driven indicator assignment method, which can effectively mine historical data information and provide an objective reference for the interval assignment of indicators; in addition, deep neural networks participate in power grid risk prediction, which can predict future operating risks from current operating risk status with higher accuracy, and provide convenience for forward-looking scheduling and situation awareness of power grids.
[0139] 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 method for predicting the risk of power grid sequential operation, characterized in that: include: Acquire power grid operation parameters, and construct an AC power grid optimization dispatching model based on second-order cone relaxation according to the power grid operation parameters, wherein the AC power grid optimization dispatching model takes the minimum overall operation cost as an objective function and takes into account relevant constraints of power grid operation; Based on the AC power grid optimization dispatching model, a risk assessment system is constructed by considering four indicators, namely, frequency margin, capacity reserve, line overload and voltage deviation, and the risk assessment system is used to assign values to the power grid operation risks; A deep neural network training model is constructed using the risk assessment data of the assigned evaluation, the risk assessment data of four consecutive time sections are used as input of the deep neural network training model, and the risk assessment data of a future time section is output, so as to predict the risk situation of the future time section based on the risk assessment data of the current time section.
2. The method for predicting the risk of power grid sequential operation according to claim 1, characterized in that: The construction of the AC power grid optimization dispatching model based on second-order cone relaxation includes: According to the obtained grid operation parameters, the second-order cone relaxation of the nonlinear terms of the AC grid is adopted. The overall operation cost is minimized as the objective function. The reconstruction constraints of the nonlinear terms of the grid operation, the node power balance constraints, the branch flow constraints, the generator set operation status constraints, and the generator set output constraints are taken into account to construct an AC grid optimization scheduling model.
3. The method for predicting the risk of power grid sequential operation according to claim 2, characterized in that: The objective function of the AC power grid optimization dispatching model is expressed as: Among them, h v,t (z v,t ,u v,t ,v v,t ) represents the fixed cost of generator set v at time t, c v represents the unit power generation cost of generator v at time t, z v,t represents an integer variable, which is the operating status of the generator set v at time t, u v,t represents an integer variable, which is the starting state of the generator set v at time t, v v,t represents an integer variable, which is the shutdown state of the generator set v at time t, Represents the active output of generator set v at time t.
4. The method for predicting the risk of power grid sequential operation according to claim 1, characterized in that: The risk assessment system comprises: Considering the four indicators of frequency margin, capacity reserve, line overload and voltage deviation, they are characterized by maximum disturbance power, reserve power, line overload rate and voltage deviation rate respectively; The frequency margin is calculated as: Among them, ΔP max represents the maximum disturbance power, Δf max represents the maximum disturbance frequency, D represents the load damping constant, R represents the speed regulator adjustment constant, ω n ,ζ,α,ω r , t nadir All represent intermediate variables and have no clear physical meaning. H represents the system inertia constant, T R represents the time constant, F H represents the generator turbine parameters; The capacity reserve is calculated as: in, and They represent the upper reserve power and lower reserve power provided by the generator set v at time t, represents the total reserve power of the power grid at time t, Indicates the maximum climbing power of the generator set v, and They respectively represent the lower and upper limits of the active output of the generator set v.
5. The method for predicting the risk of power grid sequential operation according to claim 4, characterized in that: The risk assessment system construction also includes: The line overload is calculated as: β=n br / N br Where β represents the line overload rate, n br Indicates the number of lines whose transmission power exceeds 80% of the transmission capacity, N br Indicates the total number of lines; The voltage deviation is calculated as: γ=n node / N node Among them, γ represents the voltage deviation rate, n node Indicates the number of nodes whose voltage fluctuation exceeds 0.05V, N node Indicates the total number of nodes.
6. The method for predicting the risk of power grid sequential operation according to claim 5, characterized in that: The value assignment and assessment of the power grid operation risk includes: The values of each indicator under the risk assessment system under given scheduling data are normalized and then modeled using a Gaussian mixture model, which is expressed as follows: Among them, F(y) represents the Gaussian mixture model, k represents a single Gaussian component, K represents the number of Gaussian components, α k , μ k and Respectively represent the weight coefficient, expectation and variance of Gaussian component k, represents the probability density function of Gaussian component k, and y represents the normalized indicator data; Specify a confidence level and give a corresponding score table, and obtain the confidence interval based on the specified confidence level: Among them, ε represents the confidence level, δ up and δ dn Represent the upper and lower bounds of the confidence interval respectively; Each indicator under the risk assessment system is assigned a value based on the interval in which it is located: Among them, y(t) represents the indicator data at time t, E(t) = {E P (t),E R (t),E L (t),E V (t)} represents the specific values of the four indicators of frequency margin, capacity reserve, line overload and voltage deviation at time t. Represents an assignment function.
7. The method for predicting the risk of power grid sequential operation according to claim 6, characterized in that: The risk situation of future time sections predicted by the risk assessment data of the current time section includes: A risk assessment system training data is generated for the risk assessment data of each indicator after the assigned evaluation, and a deep neural network training model is constructed based on the risk assessment system training data. After the training is completed, the risk assessment data of four consecutive time sections are used as the input of the deep neural network training model, and the risk assessment data of a future time section is output to predict the risk situation of the future section based on the risk assessment data of the current section.
8. A system using the method for predicting the risk of power grid sequential operation as claimed in any one of claims 1 to 7, characterized in that: include: A parameter acquisition unit, used to acquire power grid operation parameters; A model building module, used to build an AC power grid optimization dispatching model based on second-order cone relaxation according to the power grid operation parameters, wherein the AC power grid optimization dispatching model takes the lowest overall operation cost as the objective function and takes into account the relevant constraints of the power grid operation; A value evaluation module is used to construct a risk evaluation system based on the AC power grid optimization dispatching model, taking into account four indicators: frequency margin, capacity reserve, line overload and voltage deviation, and use the risk evaluation system to perform value evaluation on power grid operation risks; The risk prediction module is used to construct a deep neural network training model through the risk assessment data of the assigned evaluation, use the risk assessment data of four consecutive time sections as the input of the deep neural network training model, output the risk assessment data of a future time section, and realize the prediction of the risk situation of the future time section by the risk assessment data of the current time section.
9. An electronic device, comprising: Memory and processor; The memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions. When the computer-executable instructions are executed by the processor, the steps of the method described in any one of claims 1 to 7 are implemented.
10. A computer-readable storage medium having computer program instructions stored thereon, wherein the computer-executable instructions, when executed by a processor, implement the steps of the method according to any one of claims 1 to 7.
Citation Information
Cited By
Dynamic risk assessment and suppression method for leakage current of photovoltaic access transformer area based on neural network and CVaR prediction algorithm
CN120952552A
