CN115829334A - Risk assessment method and system for power grid service - Google Patents
Risk assessment method and system for power grid service Download PDFInfo
- Publication number
- CN115829334A CN115829334A CN202211727197.XA CN202211727197A CN115829334A CN 115829334 A CN115829334 A CN 115829334A CN 202211727197 A CN202211727197 A CN 202211727197A CN 115829334 A CN115829334 A CN 115829334A
- Authority
- CN
- China
- Prior art keywords
- evaluation model
- risk
- value
- assessment
- model
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Images
Classifications
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y04—INFORMATION OR COMMUNICATION TECHNOLOGIES HAVING AN IMPACT ON OTHER TECHNOLOGY AREAS
- Y04S—SYSTEMS INTEGRATING TECHNOLOGIES RELATED TO POWER NETWORK OPERATION, COMMUNICATION OR INFORMATION TECHNOLOGIES FOR IMPROVING THE ELECTRICAL POWER GENERATION, TRANSMISSION, DISTRIBUTION, MANAGEMENT OR USAGE, i.e. SMART GRIDS
- Y04S10/00—Systems supporting electrical power generation, transmission or distribution
- Y04S10/50—Systems or methods supporting the power network operation or management, involving a certain degree of interaction with the load-side end user applications
Landscapes
- Supply And Distribution Of Alternating Current (AREA)
Abstract
The invention discloses a risk assessment method and a system for power grid services, wherein the method comprises the following steps: acquiring current risk index data and historical risk index data of the power grid service; constructing an initial evaluation model, and performing model training on the initial evaluation model by adopting historical risk index data to generate a target evaluation model; evaluating and predicting the current risk index data through a target evaluation model to generate a risk evaluation value; the method solves the technical problems that in the prior art, risk assessment of the power grid service is mostly carried out on a service system by engaging external experts and following a risk assessment methodology, workload is huge, assessment period is long, assessment quality is easily affected by the level of assessment experts, and the existing power grid service risk is difficult to scientifically and stably assess.
Description
Technical Field
The invention relates to the technical field of power grids, in particular to a risk assessment method and system for power grid services.
Background
The power grid industry has the characteristics of dense technical capital, instantaneous balance of supply and demand, continuous production and operation, incapability of large-scale effective storage and the like, and has higher risk. With the reform development of the power grid market, each link of power grid production is divided into four independent links of power generation, power transmission, power distribution and sale, and processes of competition among power generation enterprises, transaction between the power generation enterprises and the power grid enterprises, transaction between the power grid enterprises and power grid customers and the like are added, and characteristics of power grid products inevitably bring more risks to power grid companies. With the continuous expansion of the power grid market, the gradual enhancement of power supply supervision by government departments and the improvement of the requirements of power consumers on power supply service, the key point of power grid operation is gradually changed from a power generation side to a power selling side, the importance of power grid services is increasingly important, the security risk of the power grid services is also increasingly emphasized, and the method is an important task related to the development of power grids. In the power grid risk management process, if the risk processing is not good, the operation development of a power grid company is seriously influenced, so that the risk existing in the power grid business process needs to be evaluated, and a basis is provided for the power grid company to make a decision.
In the prior art, most of the risk assessment of the power grid service is carried out on a service system by engaging external experts and following a risk assessment methodology, so that the workload is huge, the assessment period is long, and the assessment quality is easily influenced by the level of assessment experts, so that the risk of the existing power grid service is difficult to be assessed scientifically and stably.
Disclosure of Invention
The invention provides a risk assessment method and system for power grid services, and solves the technical problems that in the prior art, risk assessment of the power grid services is carried out on a service system mostly by engaging external experts and following a risk assessment methodology, the workload is huge, the assessment period is long, the assessment quality is easily influenced by the level of assessment experts, and the existing power grid service risk is difficult to scientifically and stably assess.
The invention provides a risk assessment method of power grid service, which comprises the following steps:
acquiring current risk index data and historical risk index data of the power grid service;
constructing an initial evaluation model, and performing model training on the initial evaluation model by adopting the historical risk index data to generate a target evaluation model;
evaluating and predicting the current risk index data through the target evaluation model to generate a risk evaluation value;
and determining the risk level of the current risk index data according to the risk assessment value, generating a risk assessment strategy and outputting the risk assessment strategy.
Optionally, the step of constructing an initial evaluation model, performing model training on the initial evaluation model by using the historical risk indicator data, and generating a target evaluation model includes:
constructing an initial evaluation model, and dividing the historical risk index data into training data and testing data;
inputting the training data into the initial evaluation model, and determining a loss function value; if the loss function value is converged, generating a secondary evaluation model;
testing the secondary evaluation model by adopting the test data, and calculating the accuracy of the secondary evaluation model; and if the accuracy is greater than or equal to the accuracy threshold, generating a target evaluation model.
Optionally, the training data is input to the initial evaluation model, and a loss function value is determined; if the loss function value is converged, generating a secondary evaluation model, comprising the following steps of:
inputting the training data into the initial evaluation model, and determining a loss function value;
and if the loss function value is not converged, iteratively updating the weight parameters of the initial evaluation model in a back propagation mode until the loss function value is converged to generate a secondary evaluation model.
Optionally, the secondary evaluation model is tested by using the test data, and the accuracy of the secondary evaluation model is calculated; if the accuracy is greater than or equal to the accuracy threshold, generating a target evaluation model, including:
testing the secondary evaluation model by adopting the test data, outputting a test risk evaluation value, and determining the number of hidden layer nodes and the number of iterations of the secondary evaluation model;
calculating the accuracy of the secondary evaluation model according to the test risk evaluation value, the number of nodes of the hidden layer and the number of iterations;
if the accuracy is smaller than an accuracy threshold, performing model optimization on the secondary evaluation model by adjusting the number of hidden layer nodes and the number of iterations to generate a target evaluation model;
and if the accuracy is greater than or equal to the accuracy threshold, selecting the secondary evaluation model as a target evaluation model.
Optionally, the step of calculating an accuracy of the secondary evaluation model according to the test risk assessment value, the number of hidden layer nodes, and the number of iterations includes:
calculating a node regulation function value according to the test risk evaluation value and the number of nodes of the hidden layer;
calculating an iteration adjusting value according to the test risk assessment value and the iteration times;
and calculating the accuracy of the secondary evaluation model according to the node adjustment function value and the iteration adjustment value.
Optionally, the calculation formula of the node adjustment function value is
In the formula: beta is the node adjustment coefficient of the secondary evaluation model, a is the number of hidden layer nodes of the secondary evaluation model, P i Test Risk assessment value for Secondary assessment model, w (P) i ) Adjusting function values for nodes of the secondary evaluation model;
the iterative adjustment value is calculated by the formula
In the formula: lambda is the iterative adjustment coefficient of the secondary evaluation model, n is the iteration number of the secondary evaluation model, P i Test Risk assessment value for Secondary assessment model, f (P) i ) Iteratively adjusting the value for the secondary evaluation model;
the accuracy is calculated by the formula
In the formula: r i For the accuracy of the secondary evaluation model, a is the number of hidden layer nodes of the secondary evaluation model, n is the number of iterations of the secondary evaluation model, and f (P) i ) Iterative adjustment of the value for the two-stage evaluation model, w ′ (P i ) Adjusting the mean value of the function for the nodes of the secondary evaluation model; wherein, w ′ (P i )＝w(P i )/a。
Optionally, if the accuracy is smaller than an accuracy threshold, performing model optimization on the secondary evaluation model by adjusting the number of hidden layer nodes and the number of iterations to generate a target evaluation model, including:
if the accuracy is smaller than an accuracy threshold, sequentially adding the hidden layer nodes together to calculate a related node adjusting function value until the calculation times of the related node adjusting function value reach the node threshold;
calculating a corresponding node adjusting function average value according to all the node adjusting function values, and selecting a hidden layer node number corresponding to the node adjusting function value closest to the node adjusting function average value as a target hidden layer node number;
sequentially adding the iteration times and combining the target hidden layer node number, respectively calculating the accuracy, and selecting the iteration time corresponding to the maximum accuracy as the target iteration time until the calculation time of the accuracy reaches an iteration threshold;
and optimizing the secondary evaluation model by adopting the target hidden layer node number and the target iteration number to generate a target evaluation model.
Optionally, the step of determining a risk level of the current risk indicator data according to the risk assessment value, generating a risk assessment policy, and outputting the risk assessment policy includes:
establishing a safety risk classification level of the power grid service;
and if the risk assessment value is within a risk acceptance interval, matching the risk assessment value with the safety risk classification level, generating a risk assessment strategy and outputting the risk assessment strategy.
The invention also provides a risk assessment system of the power grid service, which comprises the following components:
the acquisition module is used for acquiring current risk index data and historical risk index data of the power grid service;
the target evaluation model generation module is used for constructing an initial evaluation model, performing model training on the initial evaluation model by adopting the historical risk index data and generating a target evaluation model;
a risk assessment value generation module, configured to perform assessment prediction on the current risk indicator data through the target assessment model to generate a risk assessment value;
and the output module is used for determining the risk level of the current risk index data according to the risk assessment value, generating a risk assessment strategy and outputting the risk assessment strategy.
Optionally, the target evaluation model generation module includes:
the construction submodule is used for constructing an initial evaluation model and dividing the historical risk index data into training data and testing data;
the training submodule is used for inputting the training data into the initial evaluation model and determining a loss function value; if the loss function value is converged, generating a secondary evaluation model;
the test sub-module is used for testing the secondary evaluation model by adopting the test data and calculating the accuracy of the secondary evaluation model; and if the accuracy is greater than or equal to the accuracy threshold, generating a target evaluation model.
According to the technical scheme, the invention has the following advantages:
acquiring current risk index data and historical risk index data of the power grid service; constructing an initial evaluation model, and performing model training on the initial evaluation model by adopting historical risk index data to generate a target evaluation model; evaluating and predicting the current risk index data through a target evaluation model to generate a risk evaluation value; the method solves the technical problems that in the prior art, risk assessment of the power grid service is mostly carried out on a service system by engaging external experts and following a risk assessment methodology, workload is huge, assessment period is long, assessment quality is easily affected by the level of assessment experts, and the existing power grid service risk is difficult to scientifically and stably assess.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly described below, and it is obvious that the drawings in the following description are only some embodiments of the present invention, and for those skilled in the art, other drawings can be obtained according to these drawings without inventive exercise.
Fig. 1 is a flowchart illustrating steps of a risk assessment method for a power grid service according to an embodiment of the present invention;
fig. 2 is a flowchart illustrating steps of a method for risk assessment of grid services according to an alternative embodiment of the present invention;
fig. 3 is a structural block diagram of a risk assessment system for power grid services according to an embodiment of the present invention.
Detailed Description
The embodiment of the invention provides a risk assessment method and system for power grid services, which are used for solving the technical problems that in the prior art, risk assessment of the power grid services is carried out on a service system mostly by engaging external experts and following a risk assessment methodology, the workload is huge, the assessment period is long, the assessment quality is easily influenced by the level of assessment experts, and the existing power grid service risk is difficult to scientifically and stably assess.
In order to make the objects, features and advantages of the present invention more obvious and understandable, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention, and it is obvious that the embodiments described below are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
Referring to fig. 1, fig. 1 is a flowchart illustrating steps of a risk assessment method for a power grid service according to an embodiment of the present invention.
The invention provides a risk assessment method for power grid services, which comprises the following steps:
In the embodiment of the application, historical risk index data of the power grid service provides a data basis for training an initial evaluation model; after the current risk index data and the historical risk index data of the power grid service are obtained, normalization processing and principal component analysis are required to be carried out on the current risk index data and the historical risk index data so as to remove redundant indexes;
wherein, the principal component analysis can be processed by SPSS software; the current risk index data can be divided into market risk, operation risk, electricity charge safety risk, power supply service risk and the like; meanwhile, the market risk is based on consumption groups, market information, country-related policy changes and the like, such as loan policy changes; the operation risk is based on the potential safety hazard of user information with accumulation effect, electricity inspection and the like; the electric charge safety risk is based on that the electric charge management period does not carry out meter reading and checking and electricity collection according to the required specification, and the electric power company does not seriously execute the electric charge dispute between the electric power users and the electric power company caused by the national electric charge policy, and the like; the power supply service risk is based on complaints of the electricity consumers.
And 102, constructing an initial evaluation model, and performing model training on the initial evaluation model by adopting historical risk index data to generate a target evaluation model.
In the embodiment of the application, after the historical risk index data is obtained, an initial evaluation model is constructed, and then the historical risk index data is adopted to train the initial evaluation model, so that a target evaluation model is generated.
And 103, evaluating and predicting the current risk index data through the target evaluation model to generate a risk evaluation value.
In the embodiment of the application, after the current risk index data is obtained, the current risk index data is evaluated and predicted through the target evaluation model, and a corresponding risk evaluation value is generated.
And 104, determining the risk level of the current risk index data according to the risk assessment value, generating a risk assessment strategy and outputting the risk assessment strategy.
In the embodiment of the application, after the risk assessment value is generated, the risk level of the current risk index data is determined, a risk assessment strategy is generated and output, and the risk assessment strategy is pushed to a risk assessment manager for business adjustment.
In the embodiment of the application, current risk index data and historical risk index data of the power grid service are obtained firstly; constructing an initial evaluation model, and performing model training on the initial evaluation model by adopting historical risk index data to generate a target evaluation model; evaluating and predicting the current risk index data through a target evaluation model to generate a risk evaluation value; according to the method, model optimization can be performed on a target evaluation model, risk evaluation can be performed on power grid services scientifically and stably, and accuracy of the risk evaluation is improved.
Referring to fig. 2, fig. 2 is a flowchart illustrating steps of a risk assessment method for grid services according to an alternative embodiment of the present invention.
The invention provides a risk assessment method for power grid services, which comprises the following steps:
In the embodiment of the present application, the specific implementation process of step 201 is similar to that of step 101 described above, and is not described herein again.
It should be noted that the acquired historical risk indicator data needs to be divided into training data and test data, which is convenient for subsequent training of the initial evaluation model.
It should be noted that, the training data is input to the initial evaluation model, and the loss function value is determined; if the loss function value is not converged, iteratively updating the weight parameters of the initial evaluation model in a back propagation mode until the loss function value is converged to generate a secondary evaluation model;
wherein, the calculation formula of the loss function value is as follows:
in the formula: loss is a loss function value of the initial evaluation model, n is a data volume of the training data, m is a risk state of the training data, f is a predicted risk evaluation value of the initial evaluation model, and y is a real risk evaluation value of the initial evaluation model.
In the embodiment of the present application, step 204 includes the following sub-steps:
and S1, testing the secondary evaluation model by adopting test data, outputting a test risk evaluation value, and determining the number of hidden layer nodes and the iteration times of the secondary evaluation model.
It should be noted that, in order to reduce the training time and complexity of the model as much as possible, the secondary evaluation model should select a reasonable number of hidden layer nodes, where the number of hidden layer nodes is at least one.
And S2, calculating the accuracy of the secondary evaluation model according to the test risk evaluation value, the number of nodes of the hidden layer and the iteration number.
In the embodiment of the present application, step S2 specifically includes: calculating a node regulation function value by testing the risk evaluation value and the number of nodes of the hidden layer; calculating an iteration adjusting value by testing the risk evaluation value and the iteration times; calculating the accuracy of the secondary evaluation model according to the node adjustment function value and the iteration adjustment value;
wherein the calculation formula of the node adjustment function value is
In the formula: beta is the node adjustment coefficient of the secondary evaluation model, a is the number of hidden layer nodes of the secondary evaluation model, P i Test Risk assessment value for Secondary assessment model, w (P) i ) Adjusting function values for nodes of the secondary evaluation model, wherein a is more than or equal to 1 and less than or equal to 10;
the iterative adjustment value is calculated by the formula
In the formula: lambda is the iterative adjustment coefficient of the secondary evaluation model, n is the iteration number of the secondary evaluation model, P i Test Risk assessment value for Secondary assessment model, f (P) i ) The iterative adjustment value of the secondary evaluation model is n which is more than or equal to 1 and less than or equal to 10;
of accuracy is calculated by the formula
In the formula: r i For the accuracy of the secondary evaluation model, a is the number of hidden layer nodes of the secondary evaluation model, n is the number of iterations of the secondary evaluation model, and f (P) i ) Iterative adjustment of the value for the two-stage evaluation model, w ′ (P i ) Adjusting the function average value for the nodes of the secondary evaluation model; wherein, w ′ (P i )＝w(P i )/a；1≤a≤10，1≤n≤10。
And S3, if the accuracy is smaller than an accuracy threshold, performing model optimization on the secondary evaluation model by adjusting the number of nodes of the hidden layer and the iteration times to generate a target evaluation model.
It should be noted that, if the accuracy is smaller than the accuracy threshold, the number of hidden layer nodes is sequentially added and the associated node adjustment function value is calculated until the number of times of calculating the associated node adjustment function value reaches the node number threshold; calculating a corresponding node adjusting function average value according to all the node adjusting function values, and selecting a hidden layer node number corresponding to the node adjusting function value closest to the node adjusting function average value as a target hidden layer node number; sequentially adding the iteration times and combining the number of the nodes of the target hidden layer, respectively calculating the accuracy, and selecting the iteration time corresponding to the maximum accuracy as the target iteration time until the calculation time of the accuracy reaches an iteration threshold; and optimizing the secondary evaluation model by adopting the number of the target hidden layer nodes and the target iteration number to generate a target evaluation model.
And S4, if the accuracy is greater than or equal to the accuracy threshold, selecting a secondary evaluation model as a target evaluation model.
It should be noted that, if the accuracy reaches the accuracy threshold, the secondary evaluation model is selected as the target evaluation model.
And step 205, evaluating and predicting the current risk index data through the target evaluation model to generate a risk evaluation value.
In the embodiment of the application, the current risk index data is input into the target evaluation model for evaluation and prediction, and a corresponding risk evaluation value is generated.
And step 206, determining the risk level of the current risk index data according to the risk assessment value, generating a risk assessment strategy and outputting the risk assessment strategy.
It should be noted that, for subsequent risk level matching, the security risk of the grid service needs to be divided in advance according to the "security risk management system implementation guidance opinion" of the national grid company and the "grid operation security risk management regulation" of the southern power grid, so as to quantitatively evaluate the security risk of the grid service;
establishing a safety risk classification level of the power grid service: the system comprises five levels, namely, an extremely low risk state (namely, such a safety risk has no harm), a low risk state (namely, such a safety risk has low harm, but the change of a service environment, service personnel and equipment is noticed, and the potential risk and the occurrence of an emergent risk are alerted), a low risk state (namely, such a safety risk has low loss, the consequences are not obvious, only the local damage to a power grid service main body can be realized, and necessary measures can be taken to control the risk or prevent the risk from spreading), a medium risk state (namely, such a safety risk can directly cause certain loss to a power company, and the adverse effect of a part of service links is brought, and the precaution measures need to be taken in time), and a high risk state (namely, such a safety risk once occurs, great loss is inevitably caused, the consequences are serious, and high attention is required to reduce the risk level as much as possible);
in the embodiment of the application, after the risk assessment value is obtained, the risk level and the cost need to be considered in the subsequent risk assessment process, and whether the risk assessment value is in a risk acceptance interval needs to be judged; if yes, indicating that the current risk index data needs to adopt risk improvement measures, matching the risk evaluation value with the safety risk classification level, generating and outputting a corresponding risk evaluation strategy;
wherein, the upper limit value of the risk acceptance interval is a risk unacceptable value, and the lower limit value of the risk acceptance interval is a risk acceptable value; if the risk assessment value is larger than the risk unacceptable value, indicating that the risk brought by rejecting the current risk index data is present; and if the risk assessment value is smaller than the risk acceptable value, indicating the risk brought by receiving the current risk index data, and taking no risk improvement measures.
In the embodiment of the application, current risk index data and historical risk index data of the power grid service are obtained firstly; constructing an initial evaluation model, and dividing historical risk index data into training data and testing data; inputting training data into an initial evaluation model, and determining a loss function value; if the loss function value is converged, generating a secondary evaluation model; testing the secondary evaluation model by adopting the test data, and calculating the accuracy of the secondary evaluation model; if the accuracy rate is greater than or equal to the accuracy threshold value, generating a target evaluation model; evaluating and predicting the current risk index data through a target evaluation model to generate a risk evaluation value; according to the method, model optimization can be performed on a target evaluation model, risk evaluation can be performed on power grid services scientifically and stably, and accuracy of the risk evaluation is improved.
Referring to fig. 3, fig. 3 is a block diagram of a risk assessment system for grid services according to an embodiment of the present invention.
The invention also provides a risk assessment system of the power grid service, which comprises the following components:
the obtaining module 301 is configured to obtain current risk indicator data and historical risk indicator data of the power grid service.
And the target evaluation model generation module 302 is configured to construct an initial evaluation model, and perform model training on the initial evaluation model by using historical risk index data to generate a target evaluation model.
And a risk assessment value generation module 303, configured to perform assessment prediction on the current risk indicator data through the target assessment model, and generate a risk assessment value.
And the output module 304 is configured to determine a risk level of the current risk index data according to the risk assessment value, generate a risk assessment policy, and output the risk assessment policy.
Optionally, the target evaluation model generation module 302 includes:
the construction submodule is used for constructing an initial evaluation model and dividing historical risk index data into training data and testing data;
the training submodule is used for inputting training data into the initial evaluation model and determining a loss function value; if the loss function value is converged, generating a secondary evaluation model;
the test sub-module is used for testing the secondary evaluation model by adopting test data and calculating the accuracy of the secondary evaluation model; and if the accuracy is greater than or equal to the accuracy threshold, generating a target evaluation model.
Optionally, the training submodule is further configured to:
inputting training data into an initial evaluation model, and determining a loss function value;
and if the loss function value is not converged, iteratively updating the weight parameters of the initial evaluation model in a back propagation mode until the loss function value is converged to generate a secondary evaluation model.
Optionally, the test sub-module is further configured to:
testing the secondary evaluation model by adopting test data, outputting a test risk evaluation value, and determining the number of nodes and the number of iterations of a hidden layer of the secondary evaluation model;
calculating the accuracy of the secondary evaluation model according to the test risk evaluation value, the number of nodes of the hidden layer and the number of iterations;
if the accuracy is smaller than the accuracy threshold, performing model optimization on the secondary evaluation model by adjusting the number of hidden layer nodes and the number of iterations to generate a target evaluation model;
and if the accuracy is greater than or equal to the accuracy threshold, selecting the secondary evaluation model as the target evaluation model.
Optionally, the output module 304 includes:
the level construction submodule is used for constructing the safety risk classification level of the power grid service;
and the output sub-module is used for matching the risk assessment value with the safety risk classification level if the risk assessment value is in the risk acceptance interval, generating a risk assessment strategy and outputting the risk assessment strategy.
In the embodiment of the application, the current risk index data and the historical risk index data of the power grid service are acquired through an acquisition module; constructing an initial evaluation model through a target evaluation model generation module, and performing model training on the initial evaluation model by adopting historical risk index data to generate a target evaluation model; evaluating and predicting the current risk index data through a target evaluation model through a risk evaluation value generation module to generate a risk evaluation value; according to the method, model optimization can be performed on a target evaluation model, risk evaluation can be performed on power grid services scientifically and stably, and accuracy of the risk evaluation is improved.
It can be clearly understood by those skilled in the art that, for convenience and simplicity of description, the specific working processes of the above-described systems, apparatuses and units may refer to the corresponding processes in the foregoing method embodiments, and are not described herein again.
In the several embodiments provided in the present application, it should be understood that the disclosed system and method may be implemented in other ways. In addition, functional units in the embodiments of the present invention may be integrated into one processing unit, or each unit may exist alone physically, or two or more units are integrated into one unit. The integrated unit can be realized in a form of hardware, and can also be realized in a form of a software functional unit.
The above embodiments are only used to illustrate the technical solution of the present invention, and not to limit the same; although the present invention has been described in detail with reference to the foregoing embodiments, it will be understood by those of ordinary skill in the art that: the technical solutions described in the foregoing embodiments may still be modified, or some technical features may be equivalently replaced; and such modifications or substitutions do not depart from the spirit and scope of the corresponding technical solutions of the embodiments of the present invention.
Claims (10)
1. A risk assessment method for grid services, the method comprising:
acquiring current risk index data and historical risk index data of the power grid service;
constructing an initial evaluation model, and performing model training on the initial evaluation model by adopting the historical risk index data to generate a target evaluation model;
evaluating and predicting the current risk index data through the target evaluation model to generate a risk evaluation value;
and determining the risk level of the current risk index data according to the risk assessment value, generating a risk assessment strategy and outputting the risk assessment strategy.
2. The method for risk assessment of power grid services according to claim 1, wherein the step of constructing an initial assessment model, performing model training on the initial assessment model by using the historical risk indicator data, and generating a target assessment model comprises:
constructing an initial evaluation model, and dividing the historical risk index data into training data and testing data;
inputting the training data into the initial evaluation model, and determining a loss function value; if the loss function value is converged, generating a secondary evaluation model;
testing the secondary evaluation model by adopting the test data, and calculating the accuracy of the secondary evaluation model; and if the accuracy is greater than or equal to the accuracy threshold, generating a target evaluation model.
3. The method for risk assessment of grid services according to claim 2, wherein said training data is input to said initial assessment model, and a loss function value is determined; if the loss function value is converged, generating a secondary evaluation model, comprising the following steps of:
inputting the training data into the initial evaluation model, and determining a loss function value;
and if the loss function value is not converged, iteratively updating the weight parameters of the initial evaluation model in a back propagation mode until the loss function value is converged to generate a secondary evaluation model.
4. The method according to claim 2, wherein the secondary evaluation model is tested using the test data and the accuracy of the secondary evaluation model is calculated; if the accuracy is greater than or equal to the accuracy threshold, generating a target evaluation model, including:
testing the secondary evaluation model by adopting the test data, outputting a test risk evaluation value, and determining the number of nodes and the number of iterations of a hidden layer of the secondary evaluation model;
calculating the accuracy of the secondary evaluation model according to the test risk evaluation value, the number of nodes of the hidden layer and the number of iterations;
if the accuracy is smaller than an accuracy threshold, performing model optimization on the secondary evaluation model by adjusting the number of hidden layer nodes and the number of iterations to generate a target evaluation model;
and if the accuracy is greater than or equal to the accuracy threshold, selecting the secondary evaluation model as a target evaluation model.
5. The method for risk assessment of grid services according to claim 4, wherein the step of calculating the accuracy of the secondary assessment model according to the test risk assessment value, the number of hidden layer nodes and the number of iterations comprises:
calculating a node regulation function value according to the test risk assessment value and the number of nodes of the hidden layer;
calculating an iteration adjustment value according to the test risk evaluation value and the iteration times;
and calculating the accuracy of the secondary evaluation model according to the node adjustment function value and the iteration adjustment value.
6. The method according to claim 5, wherein the node adjustment function value is calculated by the formula
In the formula: beta is the node adjustment coefficient of the secondary evaluation model, a is the number of hidden layer nodes of the secondary evaluation model, P i Test Risk assessment value for Secondary assessment model, w (P) i ) Adjusting function values for nodes of the secondary evaluation model;
the iterative adjustment value is calculated by the formula
In the formula: lambda is the iterative adjustment coefficient of the secondary evaluation model, n is the iteration number of the secondary evaluation model, P i Test Risk assessment value for Secondary assessment model, f (P) i ) Iteratively adjusting the value for the secondary evaluation model;
the accuracy is calculated by the formula
In the formula: r i For the accuracy of the secondary evaluation model, a is the number of hidden layer nodes of the secondary evaluation model, n is the number of iterations of the secondary evaluation model, and f (P) i ) Iterative adjustment of the value for the two-stage evaluation model, w ′ (P i ) Adjusting the mean value of the function for the nodes of the secondary evaluation model; wherein, w ′ (P i )＝w(P i )/a。
7. The risk assessment method for grid services according to claim 6, wherein if the accuracy is smaller than an accuracy threshold, the step of performing model optimization on the secondary assessment model by adjusting the number of hidden layer nodes and the number of iterations to generate a target assessment model comprises:
if the accuracy is smaller than an accuracy threshold, sequentially adding the hidden layer nodes together to calculate a related node adjusting function value until the calculation times of the related node adjusting function value reach the node threshold;
calculating a corresponding node adjusting function average value according to all the node adjusting function values, and selecting a hidden layer node number corresponding to the node adjusting function value closest to the node adjusting function average value as a target hidden layer node number;
sequentially adding the iteration times and combining the target hidden layer node number, respectively calculating the accuracy, and selecting the iteration time corresponding to the maximum accuracy as the target iteration time until the calculation time of the accuracy reaches an iteration threshold;
and optimizing the secondary evaluation model by adopting the target hidden layer node number and the target iteration number to generate a target evaluation model.
8. The risk assessment method for grid services according to claim 1, wherein the step of determining the risk level of the current risk indicator data according to the risk assessment value, generating a risk assessment policy, and outputting the risk assessment policy comprises:
establishing a safety risk classification level of the power grid service;
and if the risk assessment value is within a risk acceptance interval, matching the risk assessment value with the safety risk classification level, generating a risk assessment strategy and outputting the risk assessment strategy.
9. A risk assessment system for grid services, comprising:
the acquisition module is used for acquiring current risk index data and historical risk index data of the power grid service;
the target evaluation model generation module is used for constructing an initial evaluation model, and performing model training on the initial evaluation model by adopting the historical risk index data to generate a target evaluation model;
a risk assessment value generation module, configured to perform assessment prediction on the current risk indicator data through the target assessment model to generate a risk assessment value;
and the output module is used for determining the risk level of the current risk index data according to the risk assessment value, generating a risk assessment strategy and outputting the risk assessment strategy.
10. The risk assessment system of a grid service according to claim 9, wherein the target assessment model generation module comprises:
the construction submodule is used for constructing an initial evaluation model and dividing the historical risk index data into training data and testing data;
the training submodule is used for inputting the training data into the initial evaluation model and determining a loss function value; if the loss function value is converged, generating a secondary evaluation model;
the test sub-module is used for testing the secondary evaluation model by adopting the test data and calculating the accuracy of the secondary evaluation model; and if the accuracy is greater than or equal to the accuracy threshold, generating a target evaluation model.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211727197.XA CN115829334A (en) | 2022-12-30 | 2022-12-30 | Risk assessment method and system for power grid service |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211727197.XA CN115829334A (en) | 2022-12-30 | 2022-12-30 | Risk assessment method and system for power grid service |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN115829334A true CN115829334A (en) | 2023-03-21 |
Family
ID=85519712
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202211727197.XA Pending CN115829334A (en) | 2022-12-30 | 2022-12-30 | Risk assessment method and system for power grid service |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115829334A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117811842A (en) * | 2024-02-29 | 2024-04-02 | 南京邮电大学 | Power grid security risk assessment method based on privacy calculation |
-
2022
- 2022-12-30 CN CN202211727197.XA patent/CN115829334A/en active Pending
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117811842A (en) * | 2024-02-29 | 2024-04-02 | 南京邮电大学 | Power grid security risk assessment method based on privacy calculation |
| CN117811842B (en) * | 2024-02-29 | 2024-05-14 | 南京邮电大学 | Power grid security risk assessment method based on privacy calculation |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Yang et al. | A graph-based method for vulnerability analysis of renewable energy integrated power systems to cascading failures | |
| CN110909983B (en) | Multidimensional assessment method for electric energy quality of active power distribution network | |
| CN105930976B (en) | Node voltage sag severity comprehensive evaluation method based on weighted ideal point method | |
| CN106600458B (en) | Power grid risk comprehensive evaluation method based on improved power betweenness | |
| Du et al. | Fast cascading outage screening based on deep convolutional neural network and depth-first search | |
| CN110310031A (en) | A kind of power distribution network multidimensional methods of risk assessment | |
| EP2863509A1 (en) | System and method for analyzing oscillatory stability in electrical power transmission systems | |
| CN105303331A (en) | Transformer repair risk decision-making method | |
| CN111695830A (en) | Power resource allocation method, system and equipment | |
| Boumaiza et al. | Modeling multi-criteria decision analysis in residential PV adoption | |
| CN115829334A (en) | Risk assessment method and system for power grid service | |
| CN114202174A (en) | Electricity price risk grade early warning method and device and storage medium | |
| CN117674119A (en) | Power grid operation risk assessment method, device, computer equipment and storage medium | |
| CN112800674A (en) | Power system element reliability parameter calculation method based on extreme learning machine | |
| CN113300373A (en) | Stability margin value prediction method and device based on PRMSE evaluation index | |
| Wenjie et al. | A multi-index evaluation method of voltage sag based on the comprehensive weight | |
| CN113919162B (en) | Voltage sag risk early warning method based on simulation and multi-source measured data fusion | |
| Pan et al. | Comprehensive Evaluation of Node Voltage Sag Severity Considering the Power Grid Side and the User Side | |
| CN115796665A (en) | Multi-index carbon efficiency grading evaluation method and device for green energy power generation project | |
| Reis et al. | A fast and accurate sampler built on Bayesian inference and optimized Hamiltonian Monte Carlo for voltage sag assessment in power systems | |
| Hu et al. | Evaluation on Arrear Electricity Charges Based on Multi-attribute Group Decision Making Considering User Payment and Abnormal Electricity Consumption | |
| Li et al. | Risk group assessment based on the demands of important consumers with standby electric source | |
| Miao et al. | An innovative Bayesian sequential censored sampling inspection method and application to test design | |
| Shen et al. | Quantitative evaluation method for software quality of smart electricity meter based on analytic hierarchy process | |
| Xu et al. | Application of Data Mining Algorithms in Power Marketing Predictive Analytics |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |