Model-data combined driven power system real-time operation risk assessment method
Through the combined model-data method, the Monte Carlo method and the optimal load cutting and wind and light-blind optimal flow method are used to form a real-time grid operation scenario, and feature extraction and real-time risk assessment are combined with the convolutional neural network and support vector machine model, which solves the problem of difficulty in dealing with new energy uncertainty and volatility in traditional power systems, and achieves a more accurate assessment of the operation safety and stability of power systems.
Patent Information
- Application Number
- CN202411795595.4
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-12-09
- Publication Date
- 2025-03-18
- Estimated Expiration
- 2044-12-09
AI Technical Summary
The risk assessment methods of traditional power systems are difficult to effectively deal with the uncertainty and volatility brought about by new energy. Especially in the case of large-scale access to new energy, how to comprehensively and accurately evaluate the safety and stability of the system has become a key issue that needs to be solved urgently.
The model-data joint drive method is adopted to form a real-time grid operation scenario that takes into account multi-source uncertainty through the Monte Carlo method, and risk calculation is performed by combining the optimal load cutting and wind and light abandonment optimal flow methods, and features are extracted using convolutional neural network and support vector machine model for real-time risk assessment.
This method can more accurately evaluate the operational safety and stability of the power system, improve risk identification and early warning capabilities, and take into account the economic and safety of the power system. It has important theoretical significance and extensive application prospects.
Smart Images

Figure CN119647969A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the field of power systems, and in particular to a model-data jointly driven power system real-time operation risk assessment method. Background Art
[0002] As the proportion of new energy sources gradually increases, the risks faced by the power system have also shown unprecedented complexity and challenges. New energy power generation has significant randomness, volatility and intermittency, which has a profound impact on the stability of power grid operation. Once a major failure occurs in the power system, it may cause serious economic losses and social problems. At the same time, the scale of the power system continues to expand and the structure becomes more complex, which undoubtedly increases the difficulty of power grid risk assessment. Traditional risk assessment methods are usually based on the stable output of traditional energy and the smooth changes in load demand, which are difficult to cope with the uncertainty and volatility after the introduction of new energy.
[0003] In addition, the indicators and models used for real-time risk assessment in traditional power systems have significant limitations when facing new power systems. Their computational efficiency and speed are also unsatisfactory in supporting large-scale real-time operational risk assessment. In particular, in the context of large-scale access to new energy, how to comprehensively and accurately evaluate the safety and stability of the system has become a key issue that needs to be solved urgently. Summary of the invention
[0004] In order to solve these problems, the present invention proposes a model-data jointly driven real-time operation risk assessment method for power systems. The purpose is to solve the problem that traditional risk assessment methods cannot effectively deal with the uncertainty and volatility brought by new energy under the background of a high proportion of new energy access in the current power system. This method can not only overcome the shortcomings of traditional assessment methods, but also better deal with the uncertainty and volatility brought by new energy. By integrating accurate modeling and data analysis methods, it can monitor and evaluate the operating status of the power system in real time, discover potential risks in time and issue early warnings.
[0005] The technical solution adopted by the present invention is as follows: a model-data joint driven power system real-time operation risk assessment method, comprising the following steps:
[0006] S1. Using the Monte Carlo method to sample a large number of real-time grid operation scenarios that take into account multi-source uncertainties;
[0007] S2. Calculate the risk of each real-time operation scenario and accumulate risk indicators using the optimal load shedding and wind and solar power curtailment optimal power flow methods;
[0008] S3. Use convolutional neural networks to train and learn samples of various risk scenarios, extract meaningful and relevant features from the original data, and input the extracted features into the support vector machine model for regression training to achieve a significant improvement in the efficiency of real-time risk assessment.
[0009] Furthermore, the steps of using the Monte Carlo method to sample and form a large number of real-time operation scenarios of power grids considering multi-source uncertainties in step S1 are as follows:
[0010] S11, sampling the operating status of all system components to obtain a state vector s of the N-component system;
[0011] S12. Use the form of predicted value plus predicted error to sample and simulate the output uncertainty of wind power, photovoltaic power and load.
[0012] Furthermore, in step S11, each component state is usually divided into two types: normal and shutdown. Assuming that the shutdown of each component is independent of each other, the component has only two states: normal operation and fault shutdown. i represents the state of element i, Q i represents the failure probability, R i is a random number in the interval [0,1] that obeys a uniform distribution. The component outage model is as follows:
[0013]
[0014] In the formula, s i =1 means the component is in operation; s i = 0 means the component is out of service. The state contains N components. The state of the system is represented by a vector s:
[0015] s=(s 1 ,…,s i ,…,s N ) (2).
[0016] Further, in step S12, the output uncertainty of wind power, photovoltaic power and load is simulated by sampling in the form of predicted value plus predicted error, and the calculation method is as follows:
[0017]
[0018] Where P(t) is the actual value; is the predicted value, which is obtained based on statistical models or physical models; ε(t) is the prediction error.
[0019] Furthermore, step S2 uses the optimal load shedding and wind and solar curtailment optimal power flow method to calculate the risk of each real-time operation scenario and accumulate risk indicators. The specific steps are as follows:
[0020] S21. Perform power flow calculation and analysis on the real-time operating status obtained by sampling to determine whether the system has power flow exceeding the limit and node voltage exceeding the limit. If an abnormal operating state occurs, perform optimal load shedding and wind and solar power abandonment power flow correction on the system;
[0021] S22. Focus on risk factors in real-time operation and establish risk indicators to more accurately perceive and quantify the risk level in the real-time operation of the power grid.
[0022] Furthermore, the power flow calculation analysis equation in step S21 is as follows:
[0023]
[0024] Where: n is the number of system nodes; P i and Q i are the injected active power and injected reactive power of node i; V i and δ i are the amplitude and phase angle of the voltage at node i; δ ij =δ i -δ j , G ij and B ij They are respectively the real and imaginary parts of the admittance in the node admittance matrix;
[0025] The optimal load shedding and wind and solar power curtailment flow model is as follows:
[0026] Objective function:
[0027]
[0028] Constraints:
[0029]
[0030] Where P L,i is the load shedding power at node i; P W,i and P S,i are the abandoned wind and abandoned optical powers at node i respectively; 1 ,λ 2 are the penalty factors for load shedding and power abandonment respectively; Δt is the dispatch interval; V and δ are the node voltage amplitude vector and phase angle vector respectively; V i is an element of V; PD i and QD i are the active and reactive power of the load at node i respectively; and are the lower and upper limits of injected active and injected reactive power of node i, respectively; and are the upward and downward climbing rates of the node i unit set respectively; T kis the current on branch k; is the limit transmission capacity on branch k; and are the lower and upper limits of the voltage amplitude of node i respectively; ND, NE, NG, L and N are the sets of load nodes, new energy station nodes, power supply nodes, all branches and all nodes in the system respectively.
[0031] Furthermore, step S22 includes the following indicators:
[0032] The flexibility of the increase is not enough to meet expectations, parameter E UFNS , refers to the expected value of the difference between the reserve provided by conventional units and the actual demand during the operation day, and its calculation method is as follows:
[0033]
[0034] Where: ΔR Ut and R Ut They represent the reserve shortage and reserve capacity of the system at time t respectively; P net,t and P net,t+1 are the net load at time t and t+1 respectively; P i,max and P i,t are the output upper limit and actual output of the node i generator set at time t respectively;
[0035] The flexibility of the downward adjustment is not enough to meet expectations, and the parameter E DFNS , refers to the expected value of the difference between the reserve provided by conventional units and the actual demand during the operation day, and its calculation method is as follows:
[0036]
[0037] Where: ΔR Dt and R Dt They represent the down-regulated reserve shortage and down-regulated reserve capacity of the system at time t respectively; P i,min is the lower limit of the output of the node i generator set at time t;
[0038] Wind curtailment expectation, parameter E WEC , calculated as follows:
[0039]
[0040] Where: p wec is the probability of wind abandonment;
[0041] Light abandonment expectation, parameter E SEC , calculated as follows:
[0042]
[0043] Where: p sec The probability of the abandoned light state appearing;
[0044] Expected power shortage, parameter E ENS , calculated as follows:
[0045]
[0046] Where: p ens is the probability of load shedding state occurring.
[0047] Furthermore, in step S3, the ReLU activation function, RMSE loss function and Adam optimization algorithm are selected, and the feature matrix is constructed based on the real-time operation status of the power system as follows:
[0048]
[0049] Where n is the number of system nodes; P L,i is the load power of node i; P R,i is the new energy power of node i; P G,i is the power of conventional unit at node i.
[0050] Another object of the present invention is to provide a model-data jointly driven power system real-time operation risk assessment system, comprising: a memory, a processor, and a computer program stored on the memory and executable on the processor, wherein the computer program, when executed by the processor, implements the above-mentioned model-data jointly driven power system real-time operation risk assessment method.
[0051] Another object of the present invention is to provide a computer-readable storage medium, on which is stored an implementation program for information transmission, and when the program is executed by a processor, it implements the above-mentioned model-data jointly driven real-time operation risk assessment method for power systems.
[0052] Advantages and beneficial effects of the present invention: The present invention aims to address the randomness, volatility and intermittency of renewable energy power generation, and combines advanced and accurate modeling and data analysis methods to enhance the risk identification and early warning capabilities of the power system in a complex operating environment. Through an innovative risk assessment model, the operational safety and stability of the power system can be accurately assessed. The present invention takes into account both the economy and safety of the power system, has important theoretical significance and broad application prospects, and can provide scientific support and decision-making basis for the safe operation of a high-proportion renewable energy power system. BRIEF DESCRIPTION OF THE DRAWINGS
[0053] Figure 1 is a flow chart of the present invention;
[0054] Figure 2 This is a flow chart of a support vector machine using the radial basis kernel function as the kernel function adopted by the present invention. DETAILED DESCRIPTION
[0055] The present invention is further described below with reference to the accompanying drawings:
[0056] Example 1
[0057] like Figure 1 As shown, a model-data joint driven power system real-time operation risk assessment method includes the following steps:
[0058] S1. Monte Carlo method is used to sample a large number of real-time operation scenarios of power grids considering multi-source uncertainties. The specific steps are as follows:
[0059] S11. Sample the operating status of all system components to obtain the state vector s of the N-component system, where each component state is usually divided into two types: normal and outage. Assume that the outage of each component is independent of each other, and the component has only two states: normal operation and fault outage. s i represents the state of element i, Q i represents the failure probability, R i is a random number in the interval [0,1] that obeys a uniform distribution. The component outage model is as follows:
[0060]
[0061] In the formula, s i =1 means the component is in operation; s i = 0 means the component is out of service. The state contains N components. The state of the system is represented by a vector s:
[0062] s=(s 1 ,…,s i ,…,s N ) (2)
[0063] S12. The output uncertainty of wind power, photovoltaic power and load is simulated by sampling in the form of predicted value plus predicted error. The calculation method is as follows:
[0064]
[0065] Where P(t) is the actual value; is the predicted value, obtained based on statistical models or physical models; ε(t) is the prediction error;
[0066] S2. Use the optimal load shedding and wind and solar curtailment optimal power flow method to calculate the risk of each real-time operation scenario and accumulate risk indicators. The specific steps are as follows:
[0067] S21. Perform power flow calculation and analysis on the real-time operating status obtained by sampling to determine whether the system has power flow exceeding the limit and node voltage exceeding the limit. If an abnormal operating state occurs, perform optimal load shedding and wind and solar power abandonment power flow correction on the system. The power flow calculation and analysis equation is as follows:
[0068]
[0069] Where: n is the number of system nodes; P i and Q i are the injected active power and injected reactive power of node i; V i and δ i are the amplitude and phase angle of the voltage at node i; δ ij =δ i -δ j , G ij and B ij They are respectively the real and imaginary parts of the admittance in the node admittance matrix;
[0070] The optimal load shedding and wind and solar power curtailment flow model is as follows:
[0071] Objective function:
[0072]
[0073] Constraints:
[0074]
[0075] Where P L,i is the load shedding power at node i; P W,i and P S,i are the abandoned wind and abandoned optical powers at node i respectively; 1 ,λ 2 are the penalty factors for load shedding and power abandonment respectively; Δt is the dispatch interval; V and δ are the node voltage amplitude vector and phase angle vector respectively; V i is an element of V; PD i and QD i are the active and reactive power of the load at node i respectively; and are the lower and upper limits of injected active and injected reactive power of node i, respectively; and are the upward and downward climbing rates of the node i unit set respectively; T k is the current on branch k; is the limit transmission capacity on branch k; and are the lower and upper limits of the voltage amplitude of node i respectively; ND, NE, NG, L and N are the sets of load nodes, new energy station nodes, power supply nodes, all branches and all nodes in the system respectively;
[0076] S22. Risk indicators are established for the key risk factors in real-time operation to more accurately perceive and quantify the risk level in the real-time operation of the power grid, including the following indicators:
[0077] The flexibility of the increase is not enough to meet expectations, parameter E UFNS , refers to the expected value of the difference between the reserve provided by conventional units and the actual demand during the operation day, and its calculation method is as follows:
[0078]
[0079] Where: ΔR Ut and R Ut They represent the reserve shortage and reserve capacity of the system at time t respectively; P net,t and P net,t+1 are the net load at time t and t+1 respectively; P i,max and P i,t are the output upper limit and actual output of the node i generator set at time t respectively;
[0080] The flexibility of the downward adjustment is not enough to meet expectations, and the parameter E DFNS , refers to the expected value of the difference between the reserve provided by conventional units and the actual demand during the operation day, and its calculation method is as follows:
[0081]
[0082] Where: ΔR Dt and R Dt They represent the down-regulated reserve shortage and down-regulated reserve capacity of the system at time t respectively; P i,min is the lower limit of the output of the node i generator set at time t;
[0083] Wind curtailment expectation, parameter E WEC , calculated as follows:
[0084]
[0085] Where: p wec is the probability of wind abandonment;
[0086] Light abandonment expectation, parameter E SEC , calculated as follows:
[0087]
[0088] Where: psec The probability of the abandoned light state appearing;
[0089] Expected power shortage, parameter E ENS , calculated as follows:
[0090]
[0091] Where: p ens is the probability of load shedding state occurring;
[0092] S3, using convolutional neural network to train and learn each risk scenario sample, extracting meaningful and relevant features from the original data, and inputting the extracted features into the support vector machine model for regression training to achieve a significant improvement in the efficiency of real-time risk assessment; wherein, the support vector machine of this embodiment selects the radial basis kernel function as the kernel function, and the specific process is as follows Figure 2 As shown, this embodiment uses the ReLU activation function, RMSE loss function and Adam optimization algorithm, and constructs the feature matrix based on the real-time operating status of the power system as follows:
[0093]
[0094] Where n is the number of system nodes; P L,i is the load power of node i; P R,i is the new energy power of node i; P G,i is the power of conventional unit at node i.
[0095] In summary, the present invention takes into account both the economy and safety of the power system, has important theoretical significance and broad application prospects, and can provide scientific support and decision-making basis for the safe operation of a high-proportion new energy power system.
[0096] The above are only embodiments of the present invention, and are not intended to limit the patent scope of the present invention. Any equivalent structure or equivalent process transformation made using the contents of the present invention specification and drawings, or directly or indirectly applied in other related technical fields, are also included in the patent protection scope of the present invention.
Claims
1. A model-data-jointly driven power system real-time operation risk assessment method, characterized in that: The following steps are involved: S1. Using the Monte Carlo method to sample a large number of real-time grid operation scenarios that take into account multi-source uncertainties; S2. Calculate the risk of each real-time operation scenario and accumulate risk indicators using the optimal load shedding and wind and solar power curtailment optimal power flow methods; S3. Use convolutional neural networks to train and learn samples of various risk scenarios, extract meaningful and relevant features from the original data, and input the extracted features into the support vector machine model for regression training to achieve a significant improvement in the efficiency of real-time risk assessment.
2. According to claim 1, a model-data joint driven power system real-time operation risk assessment method is characterized in that: In step S1, the steps of sampling a large number of real-time operation scenarios of power grids considering multi-source uncertainties using the Monte Carlo method are as follows: S11, sampling the operating status of all system components to obtain a state vector s of the N-component system; S12. Use the form of predicted value plus predicted error to sample and simulate the output uncertainty of wind power, photovoltaic power and load.
3. A model-data-jointly driven power system real-time operation risk assessment method according to claim 2, characterized in that: In step S11, each component state is usually divided into two types: normal and shutdown. Assume that the shutdown of each component is independent of each other, and the component has only two states: normal operation and fault shutdown. i represents the state of element i, Q i represents the failure probability, R i is a random number in the interval [0,1] that obeys a uniform distribution. The component outage model is as follows: In the formula, s i =1 means the component is in operation; s i = 0 means the component is out of service. The state contains N components. The state of the system is represented by a vector s: s=(s1,...,s i ,...,s N ) (2).
4. A model-data-jointly driven power system real-time operation risk assessment method according to claim 3, characterized in that: Step S12: Sampling and simulating the output uncertainty of wind power, photovoltaic power and load in the form of predicted value plus predicted error, the calculation method is as follows: Where P(t) is the actual value; is the predicted value, which is obtained based on statistical models or physical models; ε(t) is the prediction error.
5. A model-data-jointly driven power system real-time operation risk assessment method according to claim 4, characterized in that: Step S2 uses the optimal load shedding and wind and solar curtailment optimal power flow method to calculate the risk of each real-time operation scenario and accumulate risk indicators. The specific steps are as follows: S21. Perform power flow calculation and analysis on the real-time operating status obtained by sampling to determine whether the system has power flow exceeding the limit and node voltage exceeding the limit. If an abnormal operating state occurs, perform optimal load shedding and wind and solar power abandonment power flow correction on the system; S22. Focus on risk factors in real-time operation and establish risk indicators to more accurately perceive and quantify the risk level in the real-time operation of the power grid.
6. A model-data-jointly driven power system real-time operation risk assessment method according to claim 5, characterized in that: The power flow calculation analysis equation in step S21 is as follows: Where: n is the number of system nodes; P i and Q i are the injected active power and injected reactive power of node i; V i and δ i are the amplitude and phase angle of the voltage at node i; δ ij =δ i -δ j , G ij and B ij They are respectively the real and imaginary parts of the admittance in the node admittance matrix; The optimal load shedding and wind and solar power curtailment flow model is as follows: Objective function: Constraints: Where P L,i is the load shedding power at node i; P W,i and P S,i are the wind power and solar power abandonment at node i respectively; λ1 and λ2 are the penalty factors for load shedding and power abandonment respectively; Δt is the scheduling interval; V and δ are the node voltage amplitude vector and phase angle vector respectively; V i is an element of V; PD i and QD i are the active and reactive power of the load at node i respectively; and are the lower and upper limits of injected active and injected reactive power of node i, respectively; and They are the upward and downward climbing rates of the node i unit set respectively; T k is the flow on branch k; T k max is the limit transmission capacity on branch k; V i min and V i max They are the lower and upper limits of the voltage amplitude of node i respectively; ND, NE, NG, L and N are the load nodes, new energy station nodes, power supply nodes, and the collection of all branches and all nodes in the system respectively.
7. A model-data-jointly driven power system real-time operation risk assessment method according to claim 6, characterized in that: Step S22 includes the following indicators: The flexibility of the upward adjustment is not enough to meet expectations, and the parameter E UFNS , refers to the expected value of the difference between the reserve provided by conventional units and the actual demand during the operation day, and its calculation method is as follows: Where: ΔR Ut and R Ut They represent the reserve shortage and reserve capacity of the system at time t respectively; P net,t and P net,t+1 are the net load at time t and t+1 respectively; P i,max and P i,t are the output upper limit and actual output of the node i generator set at time t respectively; The flexibility of the downward adjustment is not enough to meet expectations, and the parameter E DFNS , refers to the expected value of the difference between the reserve provided by conventional units and the actual demand during the operation day, and its calculation method is as follows: Where: ΔR Dt and R Dt They represent the down-regulated reserve shortage and down-regulated reserve capacity of the system at time t respectively; P i,min is the lower limit of the output of the node i generator set at time t; Wind curtailment expectation, parameter E WEC , calculated as follows: Where: p wec is the probability of wind abandonment; Light abandonment expectation, parameter E SEC , calculated as follows: Where: p sec The probability of the abandoned light state appearing; Expected power shortage, parameter E ENS , calculated as follows: Where: p ens is the probability of load shedding state occurring.
8. The model-data-jointly driven power system real-time operation risk assessment method according to claim 7 is characterized in that: In step S3, the ReLU activation function, RMSE loss function and Adam optimization algorithm are selected, and the feature matrix is constructed based on the real-time operation status of the power system as follows: Where n is the number of system nodes; P L,i is the load power of node i; P R,i is the new energy power of node i; P G,i is the power of conventional unit at node i.
9. A model-data-jointly driven power system real-time operation risk assessment system, characterized in that: include: A memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the computer program is executed by the processor, a model-data jointly driven method for real-time operation risk assessment of a power system is implemented as described in any one of claims 1 to 8.
10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores an implementation program for information transmission, and when the program is executed by the processor, it implements a model-data jointly driven real-time operation risk assessment method for a power system as described in any one of claims 1 to 8.
Citation Information
Patent Citations
Non-intrusive load monitoring method combining unsupervised domain adaptive strategy and attention mechanism
CN113902104A
Power system flexibility quantitative evaluation method and system based on probabilistic optimal power flow
CN114970096A
Electric power system operation reliability evaluation method considering data center load flexibility
CN116384825A
Novel power system risk assessment method based on multi-head graph attention network
CN118428716A
A data collection system for use in chemical production process-related or industrial environment
CN209085657U
Cited By
Electric power system rapid static safety analysis method considering operation mode change
CN121146491A
