CN116029491B - Power dispatching management system and control method thereof - Google Patents
Power dispatching management system and control method thereof Download PDFInfo
- Publication number
- CN116029491B CN116029491B CN202211502203.1A CN202211502203A CN116029491B CN 116029491 B CN116029491 B CN 116029491B CN 202211502203 A CN202211502203 A CN 202211502203A CN 116029491 B CN116029491 B CN 116029491B
- Authority
- CN
- China
- Prior art keywords
- power grid
- data
- power
- characteristic data
- module
- 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.)
- Active
Links
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 power dispatching management system, which comprises a data acquisition module, a power dispatching management module and a power dispatching management module, wherein the data acquisition module is used for acquiring power grid operation data; the data screening module is used for screening the power grid operation data to obtain a characteristic data set; the power grid load neural network prediction module is used for predicting the near-term load and the long-term load of the power grid by using the characteristic data set; the scheduling instruction generation module is used for generating a scheduling instruction set according to the long-term load prediction result; the operation pre-command generating module is used for generating an operation pre-command set according to the scheduling instruction set and the recent load prediction result; the operation ticket generation module is used for generating a plurality of alternative operation tickets according to the operation pre-order set and the current working condition; the manual confirmation module confirms the alternative operation ticket by the dispatching staff and selects the optimal alternative operation ticket according to the actual state of the power grid.
Description
Technical Field
The invention relates to the technical field of power dispatching, in particular to a power dispatching management system and a control method thereof.
Background
The power dispatching refers to the operations of equipment maintenance and overhaul, fault treatment, balance regulation and control of the power consumption and the power transmission capacity of the power grid and the like of the power system. Because of the high complexity and relevance of power systems, power scheduling operations need to take into account various possible impact and risk factors. Along with the development of artificial intelligence technology, the automation technology is gradually integrated into power dispatching operation, and a neural network prediction model is commonly used for predicting the running load and state of a power grid, and then a dispatcher makes a corresponding dispatching instruction according to a prediction result. However, the neural network prediction model predicts based on the power grid operation data, and because the power grid operation data is huge and contains a large amount of inaccurate interference data, the operation time of the neural network prediction model in processing the power grid operation data is long, and the accuracy of the obtained prediction result is poor, so that the accuracy of a scheduling instruction sent by a scheduling personnel is directly influenced.
Disclosure of Invention
The invention aims to solve the technical problem of providing a power dispatching management system and a control method thereof, which can solve the defects of the prior art, improve the accuracy of dispatching instructions sent by dispatching personnel and reduce the workload of the dispatching personnel.
In order to solve the technical problems, the technical scheme adopted by the invention is as follows.
A power scheduling management system includes,
the data acquisition module is used for acquiring power grid operation data;
the data screening module is used for screening the power grid operation data to obtain a characteristic data set;
the power grid load neural network prediction module is used for predicting the near-term load and the long-term load of the power grid by using the characteristic data set;
the scheduling instruction generation module is used for generating a scheduling instruction set according to the long-term load prediction result;
the operation pre-command generating module is used for generating an operation pre-command set according to the scheduling instruction set and the recent load prediction result;
the operation ticket generation module is used for generating a plurality of alternative operation tickets according to the operation pre-order set and the current working condition;
and the manual confirmation module confirms the alternative operation ticket by dispatching staff, and selects the optimal alternative operation ticket as a final operation ticket to be sent to the operation terminal according to the actual state of the power grid.
The control method of the power dispatching management system comprises the following steps:
A. the data acquisition module acquires power grid operation data;
B. the data screening module screens and reconstructs the power grid operation data to obtain a characteristic data set;
C. the power grid load neural network prediction module predicts the recent load and the long-term load of the power grid by using the characteristic data set;
D. the scheduling instruction generating module generates a scheduling instruction set according to the long-term load prediction result;
E. the operation pre-instruction generating module generates an operation pre-instruction set according to the scheduling instruction set and the recent load prediction result;
F. the operation ticket generation module generates a plurality of alternative operation tickets according to the operation pre-order set and the current working condition;
G. and the dispatching staff confirms the alternative operation ticket through the manual confirmation module, and selects the optimal alternative operation ticket as the final operation ticket to be sent to the operation terminal according to the actual state of the power grid.
Preferably, in step B, the screening and reconstructing of the grid operation data comprises the steps of,
b1, grouping power grid operation data according to a geographic area;
b2, carrying out noise reduction and duplication removal treatment on each group of power grid operation data;
b3, screening the power grid operation data processed in the step B2 to obtain original characteristic data;
and B4, reconstructing the original characteristic data to obtain a characteristic data set.
Preferably, in step B2, the noise reduction and duplication removal process for each set of grid operation data includes the following steps,
b21, classifying the power grid operation data according to the data types;
b22, fitting each type of data, and marking a time period containing peak noise data in a fitting curve;
b23, calculating correlation coefficients among different fitting curves in each peak noise time period, and deleting peak noise data in the time period if the variance of all the correlation coefficients is larger than a set threshold value;
b24, setting a sliding window, traversing the fitting curves processed in the step B23, establishing a correlation function set among different fitting curves, and if all fitting curves in the sliding window are unchanged or fluctuation ranges of all fitting curves in the sliding window are smaller than a set threshold value and all correlation functions in the correlation function set are unchanged, retaining data with the closest median value between each type of data in the sliding window and the rest of data are deleted.
Preferably, in step B3, the screening of the grid operation data comprises the steps of,
and B24, the data reserved in the step B24 are counted into original characteristic data, then all the association functions in the association function set are traversed sequentially by using the sliding window in the step B24, and when the middle point of the sliding window is coincident with the standing point or the inflection point of the association function, all the power grid operation data in the time period of the sliding window are extracted to be counted into the original characteristic data.
Preferably, in step B4, reconstructing the raw feature data comprises the steps of,
the original characteristic data are grouped according to time periods, and the original characteristic data in the same time period are grouped into a group; establishing a hash table storage structure for each group of original characteristic data, wherein each storage unit in the list storage structure comprises one original characteristic data and an offset coefficient of the original characteristic data, the offset coefficient is in direct proportion to Euclidean distance between the original characteristic data and a fitting curve, each list storage structure also comprises a storage unit for storing an original characteristic data association degree matrix, each element of the association degree matrix is two original characteristic data association coefficients, and a row-column value of the element in the association degree matrix is a storage position number of the two original characteristic data in the hash table storage structure; the entire hash table storage structure established constitutes the feature data set.
Preferably, in the step C, the power grid load neural network prediction module reads and calculates the characteristic data set in a manner that,
the method comprises the steps of C1, inputting all hash table storage structures in a characteristic data set into a power grid load neural network prediction module according to time sequence;
the power grid load neural network prediction module sequentially reads a plurality of original characteristic data according to the sequence from small to large of the offset coefficient;
c3, reading a plurality of original characteristic data which are associated with the original characteristic data read in the step C2 of the round through the association degree matrix according to the sequence of the association coefficient from large to small, wherein the number of the original characteristic data read in the step C3 is equal to the number of the original characteristic data read in the step C2;
and C4, if the power grid load neural network prediction module obtains a prediction result, ending the step C, otherwise, turning to the step C2.
Preferably, in step D, generating a scheduling instruction set based on the long-term load prediction result includes the steps of,
according to the power grid maintenance plan of each area, a combination list of the maintenance time of different equipment and the synchronous maintenance among different equipment is obtained, a low-load time period corresponding to the equipment maintenance time is matched in a long-term load prediction result, then according to whether the equipment can be electrified to operate and calculate the attenuation of the power grid transmission power in the matched low-load time period in the equipment maintenance process, then according to the combination list of the synchronous maintenance among different equipment, the total attenuation of the power grid transmission power by each synchronous maintenance combination is calculated, when the transmission power standby redundancy of the power grid of the area is greater than the total attenuation of the power grid transmission power or the sum of the transmission power standby redundancy of other area power grids is greater than three times of the total attenuation of the power grid transmission power, a scheduling instruction is generated by using the synchronous maintenance combination, finally, a scheduling instruction set is generated, a priority index is given to the scheduling instruction according to the total attenuation of the power grid transmission power and the maintenance total time period in the scheduling instruction set, the total attenuation of the power grid power is inversely proportional to the priority index, and the total attenuation of the power transmission power is inversely proportional to the priority index.
Preferably, in step E, generating the operation prediction set according to the scheduling instruction set and the recent load prediction result includes the steps of,
and (3) obtaining all maintenance operation windows of each area according to a recent load prediction result, wherein the maintenance operation windows are time periods when the load of the area is lower than 75% of rated load and the standby redundancy of the transmission power of the regional power grid is greater than 20% of the load of the area, matching the maintenance operation windows with the scheduling instructions in the scheduling instruction set, and sequentially matching the scheduling instructions in descending order of the priority index from the scheduling instruction with the highest priority index in the matching process, wherein the total maintenance time of the matched scheduling instructions is less than 80% of the time of the corresponding maintenance operation windows.
Preferably, in the step F, generating a plurality of alternative operation tickets according to the operation pre-order set and the current working condition comprises the following steps,
and estimating the completion time of the operation ticket currently being executed according to the current working condition, screening and combining operation pre-orders in the operation pre-order set according to the estimated completion time of the operation ticket and the power transmission power change of the regional power grid after the completion of the operation ticket to obtain an alternative operation ticket, and adapting the influence of the execution time of the alternative operation ticket on the power transmission power of the regional power grid to the load change of the regional power grid in the same time period.
The beneficial effects brought by adopting the technical scheme are as follows: according to the invention, the basic functions of the existing power dispatching system are fully utilized and upgraded, and the prediction speed and the prediction accuracy of the neural network prediction model are greatly improved by optimizing the operation data of the power grid on the premise that the core neural network prediction model is not changed and the dispatching system is not required to be replaced integrally. The invention is especially suitable for the upgrade project aiming at the existing mature power dispatching system, the change of the existing power dispatching system is relatively less, and the stability of the upgraded and modified system is relatively higher.
Specifically, the invention improves the existing power dispatching system from two layers:
1. the invention optimizes the data input into the power grid load neural network prediction module. Firstly, fitting, denoising and de-duplication are carried out on the data, and double parameters of the fitting curve variable quantity and the correlation function are adopted as standards, so that various noise data and repeated data can be effectively removed, and the invalid data quantity input into the power grid load neural network prediction module is greatly reduced. And then, carrying out further secondary screening on data with obvious characteristics in the power grid operation data by using the correlation function to form original characteristic data, and particularly, effectively improving the screening quality of the original characteristic data by correlating the change state of the correlation function with the significance of the data characteristics. In order to enable the power grid load neural network prediction module to obtain useful characteristic data more efficiently, the invention also reconstructs the original characteristic data to form a hash table storage structure, so that the data reading speed is effectively improved, when the power grid load neural network prediction module reads the characteristic data, the characteristic data with smaller offset coefficient and the characteristic data with larger association degree are preferentially read, then prediction is carried out, and if a prediction result cannot be output, cyclic reading and prediction are carried out until the prediction result is output. As the characteristic significance is improved in the original characteristic data screening process, the cyclic process can maximally improve the utilization rate of effective characteristic data, so that data with higher characteristic significance is preferentially processed, and the output of a prediction result of the power grid load neural network prediction module is accelerated.
2. The accuracy of the prediction result of the power grid load neural network prediction module is improved through optimizing the power grid operation data, so that the generation of an operation ticket for computer assistance becomes possible. Firstly, generating all possible scheduling instructions according to a long-term load prediction result, fully considering coordination and coordination among power transmission powers of power grids in different areas in the process, so that different scheduling instructions can be obtained as much as possible, then selecting scheduling instructions meeting requirements according to a recent load prediction result, generating operation pre-instructions, optimizing a matching sequence of the scheduling instructions by using priorities of the scheduling instructions in the process, reducing influence of the generated operation pre-instructions on power transmission of the power grid, and finally screening and combining the operation pre-instructions according to current actual working conditions to obtain alternative operation tickets. Then, a dispatching staff can directly select an optimal alternative operation ticket from the alternative operation tickets as a final operation ticket, so that the workload of manually designing and compiling the operation ticket by the dispatching staff is greatly reduced, meanwhile, the accuracy of power grid load prediction is improved, and meanwhile, the human intervention is reduced, so that the probability of interference and human error of a prediction result error on dispatching is reduced, and the accuracy of a dispatching instruction sent by the dispatching staff is improved.
Drawings
Fig. 1 is a block diagram of one embodiment of the present invention.
In the figure: 1. a data acquisition module; 2. a data screening module; 3. the power grid load neural network prediction module; 4. a scheduling instruction generation module; 5. an operation pre-command generating module; 6. an operation ticket generation module; 7. and a manual confirmation module.
Detailed Description
In the following description of embodiments, for purposes of explanation and not limitation, specific details are set forth, such as particular system architectures, techniques, etc. in order to provide a thorough understanding of the embodiments of the application. It will be apparent, however, to one skilled in the art that the present application may be practiced in other embodiments that depart from these specific details. In other instances, detailed descriptions of well-known systems, devices, circuits, and methods are omitted so as not to obscure the description of the present application with unnecessary detail.
One embodiment of the present invention includes that,
the data acquisition module 1 is used for acquiring power grid operation data;
the data screening module 2 is used for screening the power grid operation data to obtain a characteristic data set;
the power grid load neural network prediction module 3 is used for predicting the near-term load and the long-term load of the power grid by using the characteristic data set;
the scheduling instruction generating module 4 is used for generating a scheduling instruction set according to the long-term load prediction result;
the operation pre-command generating module 5 is used for generating an operation pre-command set according to the scheduling instruction set and the recent load prediction result;
the operation ticket generation module 6 is used for generating a plurality of alternative operation tickets according to the operation pre-instruction set and the current working condition;
and the manual confirmation module 7 confirms the alternative operation ticket by dispatching staff, and selects the optimal alternative operation ticket as a final operation ticket to be sent to the operation terminal according to the actual state of the power grid.
The control method of the power dispatching management system comprises the following steps:
A. the data acquisition module 1 acquires power grid operation data;
B. the data screening module 2 screens and reconstructs the power grid operation data to obtain a characteristic data set;
the screening and reconstruction of the grid operation data comprises the steps of,
b1, grouping power grid operation data according to a geographic area;
b2, carrying out noise reduction and duplication removal treatment on each group of power grid operation data; the method comprises the following steps of,
b21, classifying the power grid operation data according to the data types;
b22, fitting each type of data, and marking a time period containing peak noise data in a fitting curve;
b23, calculating correlation coefficients among different fitting curves in each peak noise time period, and deleting peak noise data in the time period if the variance of all the correlation coefficients is larger than a set threshold value;
b24, setting a sliding window, traversing the fitting curves processed in the step B23, establishing a correlation function set among different fitting curves, and if all fitting curves in the sliding window are unchanged or the fluctuation range of all fitting curves in the sliding window is smaller than a set threshold value and all correlation functions in the correlation function set are unchanged, retaining the data with the closest median value between each type of data in the sliding window and the rest of data are deleted;
b3, screening the power grid operation data processed in the step B2 to obtain original characteristic data, comprising the following steps,
b24 retained data are counted into original characteristic data, then all association functions in the association function set are traversed sequentially by using the sliding window in the B24, and when the middle point of the sliding window is coincident with the standing point or the inflection point of the association function, all power grid operation data in the time period of the sliding window are extracted to be counted into the original characteristic data;
b4, reconstructing the original characteristic data to obtain a characteristic data set, comprising the following steps,
the original characteristic data are grouped according to time periods, and the original characteristic data in the same time period are grouped into a group; establishing a hash table storage structure for each group of original characteristic data, wherein each storage unit in the list storage structure comprises one original characteristic data and an offset coefficient of the original characteristic data, the offset coefficient is in direct proportion to Euclidean distance between the original characteristic data and a fitting curve, each list storage structure also comprises a storage unit for storing an original characteristic data association degree matrix, each element of the association degree matrix is two original characteristic data association coefficients, and a row-column value of the element in the association degree matrix is a storage position number of the two original characteristic data in the hash table storage structure; forming the built all hash table storage structures into a characteristic data set;
C. the power grid load neural network prediction module 3 predicts the near-term load and the long-term load of the power grid by using the characteristic data set; the power grid load neural network prediction module 3 reads and calculates the characteristic data set in such a way that,
c1, inputting each hash table storage structure in the characteristic data set into a power grid load neural network prediction module 3 according to time sequence;
the power grid load neural network prediction module 3 sequentially reads a plurality of original characteristic data according to the order of the offset coefficients from small to large;
c3, reading a plurality of original characteristic data which are associated with the original characteristic data read in the step C2 of the round through the association degree matrix according to the sequence of the association coefficient from large to small, wherein the number of the original characteristic data read in the step C3 is equal to the number of the original characteristic data read in the step C2;
c4, if the power grid load neural network prediction module 3 obtains a prediction result, ending the step C, otherwise, turning to the step C2;
D. the scheduling instruction generating module 4 generates a scheduling instruction set according to the long-term load prediction result, comprising the following steps,
according to the power grid maintenance plan of each area, obtaining a combination list of the maintenance time of different equipment and the synchronous maintenance among different equipment, matching a low-load time period corresponding to the equipment maintenance time in a long-term load prediction result, then calculating the attenuation of the power grid transmission power in the matched low-load time period according to whether the equipment can be electrified in the equipment maintenance process, calculating the total attenuation of the power grid transmission power by each synchronous maintenance combination according to the combination list of the synchronous maintenance among the different equipment, when the transmission power standby redundancy amount of the power grid of the area is greater than the total attenuation of the power grid transmission power or the sum of the transmission power standby redundancy amounts of other area power grids is greater than three times of the total attenuation of the power grid transmission power, generating a scheduling instruction by using the synchronous maintenance combination, finally generating a scheduling instruction set by using all the generated scheduling instructions, giving a priority index to the scheduling instruction according to the total attenuation of the power grid transmission power and the maintenance total time period in the scheduling instruction set, wherein the total attenuation of the power grid transmission power is inversely proportional to the priority index;
E. the operation pre-order generating module 5 generates an operation pre-order set according to the scheduling instruction set and the recent load prediction result, comprising the following steps,
obtaining all maintenance operation windows of each area according to a recent load prediction result, wherein the maintenance operation windows are time periods when the load of the area is lower than 75% of rated load and the standby redundancy of the transmission power of the regional power grid is greater than 20% of the load of the area, matching the maintenance operation windows with the scheduling instructions in the scheduling instruction set, and sequentially matching the scheduling instructions with the highest priority index in the matching process according to the descending order of the priority index, wherein the proportion of the total maintenance time length of the matched scheduling instructions to the corresponding maintenance operation window time length is less than 80%;
F. the operation ticket generating module 6 generates a plurality of alternative operation tickets according to the operation pre-order set and the current working condition, comprising the following steps,
estimating the completion time of the operation ticket currently being executed according to the current working condition, screening and combining operation pre-orders in the operation pre-order set according to the estimated completion time of the operation ticket and the power transmission power change of the local area power grid after the completion of the operation ticket to obtain an alternative operation ticket, and adapting the influence of the execution time of the alternative operation ticket on the power transmission power of the local area power grid to the load change of the local area in the same time period;
G. the dispatching staff confirms the alternative operation ticket through the manual confirmation module 7, and selects the optimal alternative operation ticket as the final operation ticket to be sent to the operation terminal according to the actual state of the power grid.
In addition, when the output prediction result of the power grid load neural network prediction module 3 is over time, the data deleted in the step B24 is restored, then the restored data is used for carrying out weighted average correction on the reserved data in the same sliding window, the weighting coefficient is inversely proportional to the time interval between the restored data and the reserved data, then the offset coefficient of the corrected reserved data is calculated again, and then the feature data set after correction is sent to the power grid load neural network prediction module 3. Of course, the recovery amount of the determined data can be flexibly and preferably selected according to the amount of the data and the prediction result of the power grid load neural network prediction module 3 in the actual running process, so that the data operation amount and the operation time are reduced as much as possible on the premise of ensuring the accuracy and the effectiveness of the prediction result.
Through an upgrade experiment on the dispatching system of the North electric company of the national network, the technical scheme of the invention operates stably and normally in the test operation stage of the upgraded dispatching system, the workload of dispatching personnel is reduced by about 20 percent on average, and no abnormality or accident from dispatching occurs in the time of 3 months of test operation.
In the description of the present invention, it should be understood that the terms "longitudinal," "transverse," "upper," "lower," "front," "rear," "left," "right," "vertical," "horizontal," "top," "bottom," "inner," "outer," and the like indicate or are based on the orientation or positional relationship shown in the drawings, merely to facilitate description of the present invention, and do not indicate or imply that the devices or elements referred to must have a particular orientation, be constructed and operated in a particular orientation, and thus should not be construed as limiting the present invention.
Claims (7)
1. A control method of a power scheduling management system includes,
the data acquisition module (1) is used for acquiring power grid operation data;
the data screening module (2) is used for screening the power grid operation data to obtain a characteristic data set;
the power grid load neural network prediction module (3) is used for predicting the near-term load and the long-term load of the power grid by using the characteristic data set;
the scheduling instruction generation module (4) is used for generating a scheduling instruction set according to the long-term load prediction result;
the operation pre-command generating module (5) is used for generating an operation pre-command set according to the scheduling instruction set and the recent load prediction result;
the operation ticket generation module (6) is used for generating a plurality of alternative operation tickets according to the operation pre-instruction set and the current working condition;
the manual confirmation module (7) confirms the alternative operation ticket through a dispatching worker, and selects the optimal alternative operation ticket as a final operation ticket to be sent to the operation terminal according to the actual state of the power grid;
the method is characterized by comprising the following steps of:
A. the data acquisition module (1) acquires power grid operation data;
B. the data screening module (2) screens and reconstructs the power grid operation data to obtain a characteristic data set; the screening and reconstruction of the grid operation data comprises the steps of,
b1, grouping power grid operation data according to a geographic area;
b2, carrying out noise reduction and duplication removal treatment on each group of power grid operation data; the noise reduction and duplication removal processing for each group of power grid operation data comprises the following steps,
b21, classifying the power grid operation data according to the data types;
b22, fitting each type of data, and marking a time period containing peak noise data in a fitting curve;
b23, calculating correlation coefficients among different fitting curves in each peak noise time period, and deleting peak noise data in the time period if the variance of all the correlation coefficients is larger than a set threshold value;
b24, setting a sliding window, traversing the fitting curves processed in the step B23, establishing a correlation function set among different fitting curves, and if all fitting curves in the sliding window are unchanged or the fluctuation range of all fitting curves in the sliding window is smaller than a set threshold value and all correlation functions in the correlation function set are unchanged, retaining the data with the closest median value between each type of data in the sliding window and the rest of data are deleted;
b3, screening the power grid operation data processed in the step B2 to obtain original characteristic data;
b4, reconstructing the original characteristic data to obtain a characteristic data set;
C. the power grid load neural network prediction module (3) predicts the recent load and the long-term load of the power grid by using the characteristic data set;
D. the scheduling instruction generating module (4) generates a scheduling instruction set according to the long-term load prediction result;
E. the operation pre-command generating module (5) generates an operation pre-command set according to the scheduling instruction set and the recent load prediction result;
F. the operation ticket generation module (6) generates a plurality of alternative operation tickets according to the operation pre-instruction set and the current working condition;
G. the dispatching staff confirms the alternative operation ticket through the manual confirmation module (7), and selects the best alternative operation ticket as the final operation ticket to be sent to the operation terminal according to the actual state of the power grid.
2. The control method of a power scheduling management system according to claim 1, characterized in that: in step B3, the screening of the grid operation data comprises the steps of,
and B24, the data reserved in the step B24 are counted into original characteristic data, then all the association functions in the association function set are traversed sequentially by using the sliding window in the step B24, and when the middle point of the sliding window is coincident with the standing point or the inflection point of the association function, all the power grid operation data in the time period of the sliding window are extracted to be counted into the original characteristic data.
3. The control method of a power scheduling management system according to claim 2, characterized in that: in step B4, reconstructing the raw feature data comprises the steps of,
the original characteristic data are grouped according to time periods, and the original characteristic data in the same time period are grouped into a group; establishing a hash table storage structure for each group of original characteristic data, wherein each storage unit in the list storage structure comprises one original characteristic data and an offset coefficient of the original characteristic data, the offset coefficient is in direct proportion to Euclidean distance between the original characteristic data and a fitting curve, each list storage structure also comprises a storage unit for storing an original characteristic data association degree matrix, each element of the association degree matrix is two original characteristic data association coefficients, and a row-column value of the element in the association degree matrix is a storage position number of the two original characteristic data in the hash table storage structure; the entire hash table storage structure established constitutes the feature data set.
4. A control method of a power scheduling management system according to claim 3, wherein: in the step C, the power grid load neural network prediction module (3) reads and calculates the characteristic data set in the following way,
c1, inputting each hash table storage structure in the characteristic data set into a power grid load neural network prediction module (3) according to time sequence;
c2, a power grid load neural network prediction module (3) sequentially reads a plurality of original characteristic data according to the order of the offset coefficients from small to large;
c3, reading a plurality of original characteristic data which are associated with the original characteristic data read in the step C2 of the round through the association degree matrix according to the sequence of the association coefficient from large to small, wherein the number of the original characteristic data read in the step C3 is equal to the number of the original characteristic data read in the step C2;
and C4, if the power grid load neural network prediction module (3) obtains a prediction result, ending the step C, otherwise, turning to the step C2.
5. The control method of a power scheduling management system according to claim 1, characterized in that: in step D, generating a scheduling instruction set based on the long-term load prediction result includes the steps of,
according to the power grid maintenance plan of each area, a combination list of the maintenance time of different equipment and the synchronous maintenance among different equipment is obtained, a low-load time period corresponding to the equipment maintenance time is matched in a long-term load prediction result, then according to whether the equipment can be electrified to operate and calculate the attenuation of the power grid transmission power in the matched low-load time period in the equipment maintenance process, then according to the combination list of the synchronous maintenance among different equipment, the total attenuation of the power grid transmission power by each synchronous maintenance combination is calculated, when the transmission power standby redundancy of the power grid of the area is greater than the total attenuation of the power grid transmission power or the sum of the transmission power standby redundancy of other area power grids is greater than three times of the total attenuation of the power grid transmission power, a scheduling instruction is generated by using the synchronous maintenance combination, finally, a scheduling instruction set is generated, a priority index is given to the scheduling instruction according to the total attenuation of the power grid transmission power and the maintenance total time period in the scheduling instruction set, the total attenuation of the power grid power is inversely proportional to the priority index, and the total attenuation of the power transmission power is inversely proportional to the priority index.
6. The control method of a power scheduling management system according to claim 5, wherein: in step E, generating an operation prediction set according to the scheduling instruction set and the recent load prediction result comprises the following steps,
and (3) obtaining all maintenance operation windows of each area according to a recent load prediction result, wherein the maintenance operation windows are time periods when the load of the area is lower than 75% of rated load and the standby redundancy of the transmission power of the regional power grid is greater than 20% of the load of the area, matching the maintenance operation windows with the scheduling instructions in the scheduling instruction set, and sequentially matching the scheduling instructions in descending order of the priority index from the scheduling instruction with the highest priority index in the matching process, wherein the total maintenance time of the matched scheduling instructions is less than 80% of the time of the corresponding maintenance operation windows.
7. The control method of a power scheduling management system according to claim 6, wherein: in the step F, generating a plurality of alternative operation tickets according to the operation pre-order set and the current working condition comprises the following steps,
and estimating the completion time of the operation ticket currently being executed according to the current working condition, screening and combining operation pre-orders in the operation pre-order set according to the estimated completion time of the operation ticket and the power transmission power change of the regional power grid after the completion of the operation ticket to obtain an alternative operation ticket, and adapting the influence of the execution time of the alternative operation ticket on the power transmission power of the regional power grid to the load change of the regional power grid in the same time period.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211502203.1A CN116029491B (en) | 2022-11-28 | 2022-11-28 | Power dispatching management system and control method thereof |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211502203.1A CN116029491B (en) | 2022-11-28 | 2022-11-28 | Power dispatching management system and control method thereof |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN116029491A CN116029491A (en) | 2023-04-28 |
| CN116029491B true CN116029491B (en) | 2023-07-21 |
Family
ID=86076695
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202211502203.1A Active CN116029491B (en) | 2022-11-28 | 2022-11-28 | Power dispatching management system and control method thereof |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116029491B (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117114362B (en) * | 2023-10-18 | 2024-02-06 | 江苏秉信科技有限公司 | Power grid operation plan scheduling method, system, terminal equipment and storage medium |
Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103065215A (en) * | 2012-12-19 | 2013-04-24 | 云南电力调度控制中心 | Power grid dispatching method |
Family Cites Families (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106997495A (en) * | 2017-04-13 | 2017-08-01 | 云南电网有限责任公司电力科学研究院 | A kind of Methods of electric load forecasting |
| CN110994613B (en) * | 2019-12-25 | 2021-05-18 | 杭州众工电力科技有限公司 | Power plant load scheduling system and scheduling method thereof |
| CN113570246A (en) * | 2021-07-27 | 2021-10-29 | 广东电网有限责任公司 | Regional power grid dispatching control system and control method |
| CN114897248A (en) * | 2022-05-18 | 2022-08-12 | 国网安徽省电力有限公司信息通信分公司 | Power grid load prediction method based on artificial intelligence |
-
2022
- 2022-11-28 CN CN202211502203.1A patent/CN116029491B/en active Active
Patent Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103065215A (en) * | 2012-12-19 | 2013-04-24 | 云南电力调度控制中心 | Power grid dispatching method |
Also Published As
| Publication number | Publication date |
|---|---|
| CN116029491A (en) | 2023-04-28 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN116029491B (en) | Power dispatching management system and control method thereof | |
| CN112085285B (en) | Bus load prediction method, device, computer equipment and storage medium | |
| CN103326348A (en) | System for improving local electric network power supply ability analysis and providing entire process online monitoring | |
| CN112990500A (en) | Transformer area line loss analysis method and system based on improved weighted gray correlation analysis | |
| CN111881961A (en) | Power distribution network fault risk grade prediction method based on data mining | |
| CN114742283A (en) | Medium-voltage line loop closing current estimation and risk assessment method based on deep learning network | |
| CN113673065B (en) | Loss reduction method for automatic reconstruction of power distribution network | |
| CN107563641B (en) | Disaster-resistant multi-scene differentiation planning method for power distribution network considering disaster preference | |
| CN112036686B (en) | Low-voltage distribution station area line loss evaluation method based on theoretical line loss interval calculation | |
| CN111799798B (en) | Method and system for improving accuracy of future state load flow calculation result | |
| CN111340257B (en) | Optimization method and system for maintenance plan of power transmission equipment based on risk analysis | |
| CN116502771B (en) | Power distribution method and system based on electric power material prediction | |
| CN108510164B (en) | Online generation method for external network equivalent model of designated area in power system | |
| CN111509715A (en) | Power distribution network technical transformation method and system considering key indexes to improve operation effect | |
| CN111628498A (en) | Multi-target power distribution network reconstruction method and device considering power distribution network reliability | |
| CN107992986B (en) | Long-term power grid planning method based on reliability index of power grid topological structure | |
| CN111915048A (en) | Novel power failure management method based on intelligent analysis decision theory technology | |
| CN111900765B (en) | Dynamic reactive power planning method and related device for wind power access system | |
| CN114759579A (en) | Power grid active power optimization control system, method and medium based on data driving | |
| CN114938003A (en) | Distribution network reactive compensation control method, device, equipment and storage medium | |
| CN113298402A (en) | Distribution network power supply reliability refining method and system based on artificial intelligence | |
| CN111950811A (en) | Regional photovoltaic power prediction method and system based on double-layer artificial neural network | |
| CN117495350B (en) | Clock operation and maintenance method, system, equipment and storage medium based on clock networking | |
| CN115618752B (en) | Power load control method and system based on Autoformer model | |
| CN113705862B (en) | Ultra-short-term new energy prediction data correction method in electric power spot market environment |
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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |