Power market clearing simulation redundancy constraint reduction method based on data driving

Abstract

The invention discloses an electricity market clearing simulation redundancy constraint reduction method based on data driving. The invention aims to improve the efficiency and accuracy of power market clearing simulation optimization. Comprising the following steps: collecting historical data of power market clearing, and generating a large number of power market operation scenes; and operating the power market clearing model in different scenes to generate corresponding clearing results, and taking the clearing results as simulation data. Historical data and simulation data are combined to jointly serve as a data set, the margin of constraint corresponding to each piece of data is calculated, and redundancy constraint is preliminarily judged. And taking the data set as input data, taking the clearing electric quantity of each unit as output data, dividing historical data into a training set and a test set, inputting the training set into the random forest model for training, and testing the precision by using the test set. And in a certain market operation environment, the final power market clearing model after deletion is obtained through operation, and a power market clearing result is obtained.

Description

Technical Field

[0001] The present invention relates to a power market clearing method, and more particularly to a data-driven power market clearing simulation redundant constraint reduction method. Background Art

[0002] With the rapid development of renewable energy, the complexity of the power market has increased significantly. Especially when dealing with multiple energy types and large-scale energy access, the operating conditions of the power system and the market clearing model face more and more constraints. The traditional power market clearing process often relies on fixed constraints to ensure the stability and security of the power grid. However, some of these constraints may not be necessary, and too many redundant constraints will increase the computational complexity and affect the efficiency and accuracy of market clearing. The commonly used redundant constraint identification methods currently rely mainly on empirical judgment, lack of systematicity and accuracy, and cannot effectively respond to dynamic changes in the market environment. As shown in the invention patent No. 202110034248.X, entitled Inter-provincial and Intra-provincial Power Market Coupling Clearing Method, System, Equipment and Medium, this method does not deal with redundant constraints, so the above problems exist. Summary of the invention

[0003] In view of the shortcomings of the prior art, the purpose of the present invention is to provide a data-driven method for reducing redundant constraints in power market clearing simulation, aiming to improve the efficiency and accuracy of the power market clearing process. The method first collects and processes historical data of the power market, including bidding information, clearing volume data and load demand of power generation units, and generates a large number of simulation scenarios. In these scenarios, the corresponding clearing results are obtained by running the power market clearing model. The simulation data and historical data are combined as a research data set, and the redundancy of each constraint is calculated. Then, by using the random forest model to train the data set, the importance of each input feature is evaluated, and the redundant constraint set is identified. Finally, by gradually deleting redundant constraints, repeatedly running the simulation model and verifying the optimization effect, a simplified power market clearing model is finally obtained.

[0004] To achieve the above object, the present invention provides the following technical solution: a data-driven power market clearing simulation redundant constraint reduction method, characterized in that it includes the following steps:

[0005] Step 1: Input the historical operation data of the power system and the historical clearing data of the power market, and perform data cleaning on the historical data input to the system, including deleting outliers and filling missing values;

[0006] Step 2: Generate massive scenarios based on the historical forecast data in the historical data cleaned in step 1, import the power market clearing model to generate the clearing results of the power market under various scenarios, combine the historical data with the simulation data into a data set, and preliminarily determine the redundant constraint set based on the clearing result data;

[0007] Step 3: Divide the data set in step 2 into a training set and a test set, use the training set to train the random forest model, and use the test set to test the accuracy of the model; get the importance ranking of different constraints, set a threshold, and determine that the constraints below the threshold are preliminary redundant constraints;

[0008] Step 4: Input the current operation scenario data of each power market into the power market clearing model, and gradually test the reliability of each redundant constraint by deleting the redundant constraints one by one, and judge whether the deletion of a redundant constraint causes the model to be unable to be effectively cleared. If so, update the constraint set, if not, do not update it, until all the redundant constraints are finally judged to be completed, and then obtain the final redundant constraints;

[0009] Step five, put the final electricity market clearing model after the deletion in step four into the market operation environment to run, obtain the electricity market clearing result, and complete the redundant constraint reduction of the electricity market clearing simulation.

[0010] As a further improvement of the present invention, the historical data of the power system in step 1 includes: network transmission restriction data, system load data and predicted output of new energy units, and the historical data of the power market includes: the bid amount, clearing amount and bid price of each participating market unit, the bid amount, clearing amount and bid price of each participating market load, and the bid amount, clearing amount and bid price of each new energy unit, wherein the clearing amount is recorded as Y h , represents the jth data of data category i, and the rest of the system data is recorded as X h , Represents the jth data of data category i.

[0011] As a further improvement of the present invention, the specific method of data cleaning in step 1 is:

[0012] Step 1, use the Z-score method to detect outliers:

[0013]

[0014] in and They are The mean and variance of Determine as an outlier; delete the identified outlier;

[0015] Step 1 and 2: Check for missing values ​​in the data;

[0016] Step 13: fill in the deleted and missing values ​​found in step 2, specifically by using known data to input the electricity market clearing model to calculate certain special missing data, and the remaining missing values ​​that cannot be calculated by the model are supplemented by using the average value;

[0017] Step 14: Pair the input data X h Normalize.

[0018]

[0019] is the normalized value.

[0020] As a further improvement of the present invention, the power market clearing model in step 1-3 is as follows:

[0021] Where F is the objective function, which means minimizing system cost or maximizing social welfare; g, w, d are the numbers of common units, new energy units, and loads respectively; G, W, D are the numbers of common units, new energy units, and loads respectively; α g ,α w ,α d They are the bidding prices of common units, new energy units and loads respectively; They are respectively the normal units, new energy units, and load clearance;

[0022] The relevant constraints are:

[0023]

[0024]

[0025] in are the bid amounts of common units, new energy units, and loads respectively; S L is the transformation matrix; They are the flow of each line and its up and down.

[0026] As a further improvement of the present invention, the historical forecast data in step 2 includes the forecast output of the new energy units of the power system. and predicted load Based on the historical forecast data, the mean and variance of the two can be calculated and recorded as and Based on these data, S scenarios can be generated by Monte Carlo sampling method. In the sth scenario, the output of the new energy unit is The load is Then, for the input data Xh Normalize:

[0027]

[0028] is the normalized value, and the imported electricity market clearing model is as follows:

[0029]

[0030] Among them, F is the objective function, which means minimizing system cost or maximizing social welfare; g, w, d are the numbers of common units, new energy units, and loads respectively; G, W, D are the numbers of common units, new energy units, and loads respectively; α g ,α w ,α d They are the bidding prices of common units, new energy units and loads respectively; They are respectively the normal units, new energy units, and load clearance;

[0031] The relevant constraints are:

[0032]

[0033]

[0034] in, are the bid amounts of common units, new energy units, and loads respectively; S L is the transformation matrix; They are the power flow of each line and its upper and lower limits.

[0035] As a further improvement of the present invention, the specific method of combining the historical data and the simulation data into a data set is: respectively running the clearing to obtain the corresponding clearing data is recorded as The total simulation data set is denoted as Then the historical data set and the simulated data set are merged to obtain the data set

[0036] For any constraint Calculation margin Judge u ij Is it greater than the threshold? It is set to 0.4. If it is greater than the threshold, the constraint is initially identified as a redundant constraint. The identified redundant constraint set is recorded as As a further improvement of the present invention, the data set in step 3 is divided into a training set and a test set, and the training set is used to train the random forest model, and the test set is used to test the accuracy of the model. The specific steps are as follows:

[0037] Step 3.1: transform the data set [X ij ,Yij ] is divided into training set and test set The ratio is 7:3;

[0038] Step 32: The training set Input the random forest model for training;

[0039] Step 33: Use the test set Make a prediction and get the predicted value At the same time, the mean square error MSE is calculated according to the following formula:

[0040]

[0041] Among them, n test is the number of tests.

[0042] As a further improvement of the present invention, in step 33, the input variables X of different types i are ij The specific method for judging importance is as follows:

[0043]

[0044] Among them, N t For input X ij The number of Importance(X i ) means for X i The importance of each X i Sort by importance and determine if Importance(X i ) is less than θ, its value is set to 0.1. i Less than Importance(X i ) then X i The related constraints are judged as redundant constraints and are recorded as

[0045] As a further improvement of the present invention, the specific method of judging whether a redundant constraint deletion causes the model to be unable to be effectively cleared in step 4 is: in a certain scenario s running value X ijs Input the electricity market clearing model M(·) to determine whether the model can be cleared normally and whether its clearing result is reasonable. If the model can be cleared normally and the result is reasonable, the constraint being judged is an actual redundant constraint, then U j (n) The constraint is deleted and updated to U j (n+1); otherwise, the judged constraint is not a redundant constraint. The redundant constraints initially identified are superimposed and deleted one by one until the power market clearing model in 95% of the scenarios can be cleared normally and the number of identified redundant constraints remains unchanged, then the identification is completed.j (N) is the identified redundant constraint.

[0046] As a further improvement of the present invention, in step 5, the final electricity market clearing model after the deletion in step 4 is put into the market operation environment for operation, and the specific method for obtaining the electricity market clearing result is to determine a certain market operation environment. Input the electricity market clearing model M(U j (N)) can be used to obtain the result of electricity market clearing Beneficial effects of the present invention:

[0047] 1) By identifying and deleting redundant constraints, the present invention significantly reduces the computational complexity of power market clearing, shortens the clearing time, improves the market response speed, and ensures the reliability and economy of power supply.

[0048] 2) The present invention makes the power market model more adaptable by identifying redundant constraints in real time, and can flexibly respond to market changes and adjust constraints in time, thereby improving the flexibility and efficiency of market operation. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] Figure 1 The present invention is a flow chart of a redundant constraint reduction method based on data-driven power market clearing simulation. DETAILED DESCRIPTION

[0050] The present invention will be further described below in detail with reference to the embodiments shown in the accompanying drawings.

[0051] Reference Figure 1 As shown, a data-driven power market clearing simulation redundancy constraint reduction method of this embodiment includes the following steps:

[0052] Step 1) Input the historical operation data of the power system and the historical clearing data of the power market, clean the historical data of the input system, delete outliers, fill in missing values, etc.

[0053] Step 2) Generate massive scenarios based on historical forecast data, import the power market clearing model to generate the clearing results of the power market under various scenarios, combine historical data with simulation data into a data set, and preliminarily determine the redundant constraint set based on the clearing result data.

[0054] Step 3) Divide the data set into a training set and a test set, use the training set to train the random forest model, and use the test set to test the accuracy of the model. Obtain the importance ranking of different constraints, set a threshold, and determine that the constraints below the threshold are preliminary redundant constraints.

[0055] Step 4) Input the data of each power market operation scenario into the power market clearing model, gradually check the reliability of each redundant constraint, determine whether the deletion of a redundant constraint causes the model to be unable to be effectively cleared, and update the constraint set. If all redundant constraints are finally judged to be completed, the final redundant constraints are obtained.

[0056] The step 1) is specifically as follows:

[0057] Input the historical operation data of the power system and the historical clearing data of the power market, clean the historical data of the input system, delete outliers, fill in missing values, etc.

[0058] The historical data of the power system includes but is not limited to: network transmission limitation data, system load data, predicted output of new energy units, etc.

[0059] The historical data of the electricity market include but are not limited to: the bid quantity, clearing quantity and bid price of each participating market unit, the bid quantity, clearing quantity and bid price of each participating market load, the bid quantity, clearing quantity and bid price of each new energy unit, etc.

[0060] The clearing amount is recorded as Y h ,in Represents the jth data of data category i.

[0061] Other system data are denoted as X h ,in Represents the jth data of data category i.

[0062] Perform data cleaning on historical data input into the system, delete outliers, fill in missing values, etc.

[0063] ① Check and handle outliers

[0064] Use the Z-score method to detect outliers:

[0065]

[0066] in and They are The mean and variance of Determined as an outlier.

[0067] The identified outliers are removed.

[0068] ② Check the missing values ​​of the data.

[0069] ③Fill in deleted and missing values.

[0070] For some special missing data, other known data can be used to input the electricity market clearing model for calculation. Other missing values ​​that cannot be calculated by the model can be supplemented by average value calculation.

[0071] The specific electricity market clearing model is:

[0072] Objective function:

[0073]

[0074] Where F is the objective function, which means minimizing system cost or maximizing social welfare; g, w, d are the numbers of common units, new energy units, and loads respectively; G, W, D are the numbers of common units, new energy units, and loads respectively; α g ,α w ,α d They are the bidding prices of common units, new energy units and loads respectively; They are respectively the ordinary units, new energy units, and load clearance volumes.

[0075] The relevant constraints are:

[0076]

[0077] in are the bid amounts of common units, new energy units, and loads respectively; S L is the transformation matrix; They are the power flow of each line and its upper and lower limits.

[0078] ④For input data X h Normalize.

[0079]

[0080] is the normalized value.

[0081] The step 2) is specifically as follows:

[0082] Generate massive scenarios based on historical forecast data, import the electricity market clearing model to generate the clearing results of the electricity market under various scenarios, and preliminarily determine the redundant constraint set based on each clearing result.

[0083] Historical forecast data includes the forecast output of new energy units in the power system and predicted load Based on the historical forecast data, the mean and variance of the two can be calculated and recorded as and Based on these data, S scenarios can be generated by Monte Carlo sampling method. In the sth scenario, the output of the new energy unit is The load is

[0084] For input data X h Normalize.

[0085]

[0086] is the normalized value.

[0087] Under different scenarios s, the corresponding power system scenario data is input into the power market clearing model.

[0088] The specific electricity market clearing model is:

[0089] Objective function:

[0090]

[0091] Where F is the objective function, which means minimizing system cost or maximizing social welfare; g, w, d are the numbers of common units, new energy units, and loads respectively; G, W, D are the numbers of common units, new energy units, and loads respectively; α g ,α w ,α d They are the bidding prices of common units, new energy units and loads respectively; They are respectively the ordinary units, new energy units, and load clearance volumes.

[0092] The relevant constraints are:

[0093]

[0094] in are the bid amounts of common units, new energy units, and loads respectively; S L is the transformation matrix; They are the power flow of each line and its upper and lower limits.

[0095] You can run the clearing separately to get the corresponding clearing data recorded as The total simulation data set is denoted as

[0096] Merge the historical data set and the simulated data set to obtain the data set

[0097] For any constraint Margin can be calculated Judge u ij Is it greater than the threshold? It is set to 0.4. If it is greater than the threshold, the constraint is initially identified as a redundant constraint. The identified redundant constraint set is recorded as

[0098] The step 3) is specifically as follows:

[0099] The historical data and simulation data sets are divided into training sets and test sets. The training sets are used to train the random forest model, and the test sets are used to test the accuracy of the model. The importance ranking of different constraints is obtained, and a threshold is set. The constraints below the threshold are determined as preliminary redundant constraints.

[0100] The data set [X ij ,Y ij ] is divided into training set and test set The ratio is 7:3.

[0101] The training set Input the random forest model for training. The specific random forest model is not described here, please refer to the manual.

[0102] Using the test set Make a prediction and get the predicted value The mean square error (MSE) can be calculated according to the following formula:

[0103]

[0104] Among them, n test is the number of tests.

[0105] The input variables (constraints) X of different types i can be ij The specific method for judging importance is as follows:

[0106]

[0107] Where N t For input X ij The number of Importance(X i ) means for X i The importance of.

[0108] For each X i Sort by importance and determine if Importance(X i ) is less than θ, its value is set to 0.1. i Less than Importance(X i ) then X i The related constraints are judged as redundant constraints and are recorded as

[0109] The step 4) is specifically as follows:

[0110] The data of various power market operation scenarios are input into the power market clearing model, and the reliability of each redundant constraint is gradually tested to determine whether the deletion of a redundant constraint will cause the model to be unable to be effectively cleared, and the constraint set is updated. If all redundant constraints are finally judged to be completed, the final redundant constraints are obtained.

[0111] The total set of redundant constraints initially obtained in step 2 and step 3 The constraints in are deleted and tested in the electricity market clearing model mentioned in step 2.

[0112] A scenario s running value X ijs Input the electricity market clearing model M(·) to determine whether the model can be cleared normally and whether its clearing result is reasonable. If the model can be cleared normally and the result is reasonable, the constraint being judged is an actual redundant constraint, then U j (n) The constraint is deleted and updated to U j (n+1); otherwise, the judged constraint is not a redundant constraint. The redundant constraints initially identified are superimposed and deleted one by one until the power market clearing model in 95% of the scenarios can be cleared normally and the number of identified redundant constraints remains unchanged, then the identification is completed. j (N) is the identified redundant constraint.

[0113] The step 5) is specifically as follows:

[0114] Under a certain market operating environment, the final electricity market clearing model after deletion is run to obtain the electricity market clearing result.

[0115] In a certain market environment Input the electricity market clearing model M(U j (N)) can be used to obtain the result of electricity market clearing

[0116] In summary, this embodiment discloses a data-driven method for reducing redundant constraints in power market clearing simulation. It aims to improve the efficiency and accuracy of power market clearing simulation optimization. It includes the following steps: collecting historical data of power market clearing, including bidding data and clearing quantity data of generators, load demand, clearing quantity of new energy units and the execution of related constraints, and generating a large number of power market operation scenarios based on historical new energy forecasts and load quantities. Run the power market clearing model in different scenarios to generate corresponding clearing results, which are used as simulation data. Combine historical and simulation data as a data set, calculate the margin of constraints corresponding to each data, and preliminarily judge redundant constraints. Take the data set as input data, the clearing quantity of each unit as output data, and divide the historical data into training set and test set. Input the training set into the random forest model for training, and use the test set to test the accuracy, so as to obtain the importance parameters of different inputs and identify preliminary redundant constraints. Input the power market operating environment information of different scenarios into the power market clearing model, simulate and test, gradually delete the redundant constraints initially identified, and update the constraint set. If the redundant constraints are finally judged to be complete, the final redundant constraints are obtained. Under a certain market operating environment, the final power market clearing model after deletion is run to obtain the power market clearing result.

[0117] The above is only a preferred embodiment of the present invention, and the protection scope of the present invention is not limited to the above embodiments. All technical solutions under the concept of the present invention belong to the protection scope of the present invention. It should be pointed out that for ordinary technicians in this technical field, some improvements and modifications without departing from the principle of the present invention should also be regarded as the protection scope of the present invention.

Claims

1. A data-driven method for reducing redundant constraints in power market clearing simulation, characterized in that: The steps include: Step 1: Input the historical operation data of the power system and the historical clearing data of the power market, and perform data cleaning on the historical data input to the system, including deleting outliers and filling missing values; Step 2: Generate massive scenarios based on the historical forecast data in the historical data cleaned in step 1, import the power market clearing model to generate the clearing results of the power market under various scenarios, combine the historical data with the simulation data into a data set, and preliminarily determine the redundant constraint set based on the clearing result data; Step 3: Divide the data set in step 2 into a training set and a test set, use the training set to train the random forest model, and use the test set to test the accuracy of the model; get the importance ranking of different constraints, set a threshold, and determine that the constraints below the threshold are preliminary redundant constraints; Step 4: Input the current operation scenario data of each power market into the power market clearing model, and gradually test the reliability of each redundant constraint by deleting the redundant constraints one by one, and judge whether the deletion of a redundant constraint causes the model to be unable to be effectively cleared. If so, update the constraint set, if not, do not update it, until all the redundant constraints are finally judged to be completed, and then obtain the final redundant constraints; Step five, put the final electricity market clearing model after the deletion in step four into the market operation environment to run, obtain the electricity market clearing result, and complete the redundant constraint reduction of the electricity market clearing simulation.

2. The data-driven power market clearing simulation redundancy constraint reduction method according to claim 1 is characterized by: The historical data of the power system in step 1 include: network transmission restriction data, system load data and predicted output of new energy units; the historical data of the power market include: the bid amount, clearing amount and bid price of each participating market unit, the bid amount, clearing amount and bid price of each participating market load, the bid amount, clearing amount and bid price of each new energy unit, where the clearing amount is recorded as Y h , represents the jth data of data category i, and the rest of the system data is recorded as X h , Represents the jth data of data category i.

3. The data-driven power market clearing simulation redundancy constraint reduction method according to claim 2 is characterized by: The specific method of data cleaning in step 1 is: Step 1, use the Z-score method to detect outliers: in and They are The mean and variance of Determine as an outlier; delete the identified outlier; Step 1 and 2: Check for missing values ​​in the data; Step 13: fill in the deleted and missing values ​​found in step 2, specifically by using known data to input the electricity market clearing model to calculate certain special missing data, and the remaining missing values ​​that cannot be calculated by the model are supplemented by using the average value; Step 14: Pair the input data X h Normalize. is the normalized value.

4. The data-driven power market clearing simulation redundancy constraint reduction method according to claim 3 is characterized by: The electricity market clearing model in step 1-3 is as follows: Where F is the objective function, which means minimizing system cost or maximizing social welfare; g, w, d are the numbers of common units, new energy units, and loads respectively; G, W, D are the numbers of common units, new energy units, and loads respectively; α g ,α w ,α d They are the bidding prices of common units, new energy units and loads respectively; They are respectively the normal units, new energy units, and load clearance; The relevant constraints are: in are the bid amounts of common units, new energy units, and loads respectively; S L is the transformation matrix; P l L , P l L They are the power flow of each line and its upper and lower limits.

5. The data-driven power market clearing simulation redundant constraint reduction method according to any one of claims 1 to 4, characterized in that: The historical forecast data in step 2 includes the forecast output of the new energy units of the power system and predicted load Based on the historical forecast data, the average value and variance of the two can be calculated and recorded as and Based on these data, S scenarios can be generated by Monte Carlo sampling method. In the sth scenario, the output of the new energy unit is The load is Then, for the input data X h Normalize: is the normalized value, and the imported electricity market clearing model is as follows: Among them, F is the objective function, which means minimizing system cost or maximizing social welfare; g, w, d are the numbers of common units, new energy units, and loads respectively; G, W, D are the numbers of common units, new energy units, and loads respectively; α g ,α w ,α d They are the bidding prices of common units, new energy units and loads respectively; They are respectively the normal units, new energy units, and load clearance; The relevant constraints are: in, are the bid amounts of common units, new energy units, and loads respectively; S L is the transformation matrix; P l L , P l L They are the power flow of each line and its upper and lower limits.

6. The data-driven power market clearing simulation redundancy constraint reduction method according to claim 5 is characterized by: The specific method of combining the historical data and the simulation data into a data set is: respectively run the clearing to obtain the corresponding clearing data and record it as The total simulation data set is denoted as Then the historical data set and the simulated data set are merged to obtain the data set For any constraint Calculation margin Judge u ij Is it greater than the threshold? It is set to 0.

4. If it is greater than the threshold, the constraint is initially identified as a redundant constraint. The identified redundant constraint set is recorded as 7. The data-driven power market clearing simulation redundant constraint reduction method according to any one of claims 1 to 4, characterized in that: In step 3, the data set is divided into a training set and a test set, and the training set is used to train the random forest model, and the test set is used to test the accuracy of the model. The specific steps are as follows: Step 3.1: transform the data set [X ij ,Y ij ] is divided into training set and test set The ratio is 7:3; Step 32: The training set Input the random forest model for training; Step 33: Use the test set Make a prediction and get the predicted value At the same time, the mean square error MSE is calculated according to the following formula: Among them, n test is the number of tests.

8. The data-driven power market clearing simulation redundancy constraint reduction method according to claim 7 is characterized by: In step 33, the input variables X of different types i are ij The specific method for judging importance is as follows: Among them, N t For input X ij The number of Importance(X i ) means for X i The importance of each X i Sort by importance and determine if Importance(X i ) is less than θ, its value is set to 0.

1. i Less than Importance(X i ) then X i The related constraints are judged as redundant constraints and are recorded as 9. The data-driven power market clearing simulation redundant constraint reduction method according to any one of claims 1 to 4, characterized in that: The specific method for determining whether the deletion of a redundant constraint causes the model to be unable to be effectively cleared in step 4 is: ijs Input the electricity market clearing model M(·) to determine whether the model can be cleared normally and whether its clearing result is reasonable. If the model can be cleared normally and the result is reasonable, the constraint being judged is an actual redundant constraint, then U j (n) The constraint is deleted and updated to U j (n+1); otherwise, the judged constraint is not a redundant constraint. The redundant constraints initially identified are superimposed and deleted one by one until the power market clearing model in 95% of the scenarios can be cleared normally and the number of identified redundant constraints remains unchanged, then the identification is completed. j (N) is the identified redundant constraint.

10. The data-driven power market clearing simulation redundancy constraint reduction method according to claim 9 is characterized in that: In step 5, the final electricity market clearing model after the deletion in step 4 is put into the market operation environment for operation. The specific method for obtaining the electricity market clearing result is to determine the Input the electricity market clearing model M(U j (N)) can be used to obtain the result of electricity market clearing

Images