CN117638882A - Loss prediction method, device, equipment, medium and product of electric power system - Google Patents
Loss prediction method, device, equipment, medium and product of electric power system Download PDFInfo
- Publication number
- CN117638882A CN117638882A CN202311565971.6A CN202311565971A CN117638882A CN 117638882 A CN117638882 A CN 117638882A CN 202311565971 A CN202311565971 A CN 202311565971A CN 117638882 A CN117638882 A CN 117638882A
- Authority
- CN
- China
- Prior art keywords
- load
- data
- value
- actual
- loss
- 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
- 238000000034 method Methods 0.000 title claims abstract description 62
- 238000012549 training Methods 0.000 claims abstract description 29
- 238000004590 computer program Methods 0.000 claims description 24
- 238000007621 cluster analysis Methods 0.000 claims description 11
- 230000008569 process Effects 0.000 description 11
- 238000004891 communication Methods 0.000 description 8
- 238000010586 diagram Methods 0.000 description 8
- 230000006870 function Effects 0.000 description 8
- 238000004458 analytical method Methods 0.000 description 7
- 230000004913 activation Effects 0.000 description 6
- 238000004422 calculation algorithm Methods 0.000 description 5
- 238000004364 calculation method Methods 0.000 description 5
- 238000012937 correction Methods 0.000 description 5
- 239000013598 vector Substances 0.000 description 4
- 230000002159 abnormal effect Effects 0.000 description 3
- 230000006978 adaptation Effects 0.000 description 3
- 230000008859 change Effects 0.000 description 3
- 238000003064 k means clustering Methods 0.000 description 3
- 238000012545 processing Methods 0.000 description 3
- 230000006399 behavior Effects 0.000 description 2
- 230000005540 biological transmission Effects 0.000 description 2
- 238000013500 data storage Methods 0.000 description 2
- 238000000605 extraction Methods 0.000 description 2
- 238000011478 gradient descent method Methods 0.000 description 2
- 230000006872 improvement Effects 0.000 description 2
- 210000002569 neuron Anatomy 0.000 description 2
- 238000010248 power generation Methods 0.000 description 2
- 238000000926 separation method Methods 0.000 description 2
- 230000003068 static effect Effects 0.000 description 2
- OKTJSMMVPCPJKN-UHFFFAOYSA-N Carbon Chemical compound [C] OKTJSMMVPCPJKN-UHFFFAOYSA-N 0.000 description 1
- 230000003321 amplification Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000001413 cellular effect Effects 0.000 description 1
- 238000007405 data analysis Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 229910021389 graphene Inorganic materials 0.000 description 1
- 238000003384 imaging method Methods 0.000 description 1
- 239000004973 liquid crystal related substance Substances 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000003199 nucleic acid amplification method Methods 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 230000000007 visual effect Effects 0.000 description 1
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 application relates to a loss prediction method, device, equipment, medium and product of a power system. The method comprises the following steps: acquiring predicted power data and actual power data of a power system; determining a target loss prediction model, correcting an output layer of an initial loss prediction model according to actual power data by the target loss prediction model, and training the initial loss prediction model according to predicted power data; and acquiring real-time load data, inputting the real-time load data into a target loss prediction model to predict the energy loss of the power system, and obtaining a real-time loss predicted value. The method can improve the adaptability and generalization capability of the model, avoid the problems of data deviation, data error, data integrity and the like, and improve the accuracy and reliability of the prediction model.
Description
Technical Field
The application relates to the technical field of power system prediction, in particular to a loss prediction method, device, equipment, medium and product of a power system.
Background
With the continuous development of the power system, the power transmission line of the power system is longer and longer, and the power load demand is also larger and larger. Because the energy loss caused by the electric wire resistance, the transformer and other power equipment exists in the transmission process, the energy loss is predicted, so that the supply and demand balance of the load of each node of the electric power system can be better realized, the stability and the reliability of the electric power system are improved, the energy utilization efficiency of the electric power system can be improved, and the energy cost of the electric power system is reduced.
At present, based on the load demand of each node in the power system, under the goal of minimizing energy loss, an optimal distribution scheme of the output power of each generator set is determined, and the energy loss value of the power system is predicted. However, the prediction method in the prior art has large error and low accuracy.
Disclosure of Invention
In view of the foregoing, it is desirable to provide a loss prediction method, apparatus, device, medium, and product for an electric power system.
In a first aspect, the present application provides a method for predicting loss of an electric power system, including:
acquiring predicted power data and actual power data of a power system;
determining a target loss prediction model, correcting an output layer of an initial loss prediction model according to actual power data by the target loss prediction model, and training the initial loss prediction model according to predicted power data;
and acquiring real-time load data, inputting the real-time load data into a target loss prediction model to predict the energy loss of the power system, and obtaining a real-time loss predicted value.
In one embodiment, wherein the predicted power data comprises a load prediction data set and a loss prediction data set; the actual power data includes an actual load data set and an actual loss data set; the target loss prediction model is obtained by correcting an output layer of the initial loss prediction model according to actual power data, and comprises the following steps:
For each actual load value of the actual load data set, calculating the distance between the actual load value and each load predicted value of the load predicted data set, and taking the load predicted value with the smallest distance as a target load predicted value;
determining an actual loss value corresponding to the actual load value in the actual loss data set;
and correcting the output of the target load predicted value in the initial loss predicted model according to the actual loss value to obtain a target loss predicted model.
In one embodiment, the initial loss prediction model is trained from predicted power data, comprising:
under each load predicted value, iteratively updating the output power of each generator set in the power system to obtain set output power data;
and determining an initial loss prediction model, wherein the initial loss prediction model is obtained by training according to the predicted power data and the unit output power data.
In one embodiment, for each actual load value of the actual load data set, calculating a distance between the actual load value and each load prediction value of the load prediction data set, and setting the load prediction value at the minimum distance as the target load prediction value includes:
performing cluster analysis on the actual load data set to obtain a plurality of target actual load values;
For each target actual load value, the distance between the target actual load value and each load predicted value is calculated, and the load predicted value at which the distance is the smallest is used as the target load predicted value.
In one embodiment, performing cluster analysis on the actual load data set to obtain a plurality of target actual load values includes:
acquiring the number of clusters, and determining initial load cluster centers of a plurality of actual load data sets according to the number of clusters;
calculating a second distance between the actual load value and each initial load clustering center according to each actual load value, and dividing the actual load values of the actual load data sets according to the second distances to obtain a plurality of actual load data subsets;
and carrying out iterative updating on the initial load clustering center according to the average value of the load actual values in each actual load data subset, and obtaining each target actual load value under the condition that the preset termination condition is met.
In one embodiment, calculating the second distance of the actual load value from each initial load cluster center includes:
calculating the difference value between the actual load value and each initial load clustering center to obtain a plurality of difference values corresponding to each actual load value;
carrying out average value operation on the square value of each difference value to obtain a mean square value;
And performing square root operation on the mean square value to obtain a second distance.
In a second aspect, the present application further provides a loss prediction apparatus of an electric power system, including:
the data acquisition module is used for acquiring predicted power data and actual power data of the power system;
the model acquisition module is used for determining a target loss prediction model, the target loss prediction model is obtained by correcting an output layer of an initial loss prediction model according to actual power data, and the initial loss prediction model is obtained by training according to predicted power data;
the loss prediction module is used for acquiring real-time load data, inputting the real-time load data into the target loss prediction model to predict the energy loss of the power system, and obtaining a real-time loss prediction value.
In a third aspect, the present application also provides a computer device comprising a memory and a processor, the memory storing a computer program, the processor implementing the following steps when executing the computer program:
acquiring predicted power data and actual power data of a power system;
determining a target loss prediction model, correcting an output layer of an initial loss prediction model according to actual power data by the target loss prediction model, and training the initial loss prediction model according to predicted power data;
And acquiring real-time load data, inputting the real-time load data into a target loss prediction model to predict the energy loss of the power system, and obtaining a real-time loss predicted value.
In a fourth aspect, the present application also provides a computer readable storage medium having stored thereon a computer program which when executed by a processor performs the steps of:
acquiring predicted power data and actual power data of a power system;
determining a target loss prediction model, correcting an output layer of an initial loss prediction model according to actual power data by the target loss prediction model, and training the initial loss prediction model according to predicted power data;
and acquiring real-time load data, inputting the real-time load data into a target loss prediction model to predict the energy loss of the power system, and obtaining a real-time loss predicted value.
In a fifth aspect, the present application also provides a computer program product comprising a computer program which, when executed by a processor, performs the steps of:
acquiring predicted power data and actual power data of a power system;
determining a target loss prediction model, correcting an output layer of an initial loss prediction model according to actual power data by the target loss prediction model, and training the initial loss prediction model according to predicted power data;
And acquiring real-time load data, inputting the real-time load data into a target loss prediction model to predict the energy loss of the power system, and obtaining a real-time loss predicted value.
The loss prediction method, the loss prediction device, the loss prediction equipment, the loss prediction medium and the loss prediction product of the electric power system are characterized by acquiring prediction electric power data and actual electric power data of the electric power system; determining a target loss prediction model, correcting an output layer of an initial loss prediction model according to actual power data by the target loss prediction model, and training the initial loss prediction model according to predicted power data; and acquiring real-time load data, inputting the real-time load data into a target loss prediction model to predict the energy loss of the power system, and obtaining a real-time loss predicted value. According to the method, the output layer parameters of the initial loss model are corrected by adopting actual power data, so that the adaptation capacity and generalization capacity of the model can be improved, the training data of the initial loss prediction model are power prediction data, the problems of data noise, abnormal values or error marks and the like possibly exist, and the problems of data deviation, data errors, data integrity and the like can be avoided by correcting the initial loss prediction model by using the actual power data, so that the accuracy and the reliability of the prediction model are improved.
Drawings
In order to more clearly illustrate the embodiments of the present application or the technical solutions in the related art, the drawings that are required to be used in the embodiments or the related technical descriptions will be briefly described below, and it is obvious that the drawings in the following description are only some embodiments of the present application, and other drawings may be obtained according to the drawings without inventive effort for a person having ordinary skill in the art.
FIG. 1 is an application environment diagram of a loss prediction method of a power system in one embodiment;
FIG. 2 is a flow chart of a method of loss prediction for an electrical power system in one embodiment;
FIG. 3 is a schematic flow diagram of model correction in one embodiment;
FIG. 4 is a schematic diagram of a Euclidean distance column in one embodiment;
FIG. 5 is a schematic flow chart of the amplification data in one embodiment;
FIG. 6 is a flow diagram of cluster analysis in one embodiment;
FIG. 7 is a flow chart of a method for predicting loss of a power system according to another embodiment;
FIG. 8 is a block diagram of a loss prediction apparatus of a power system in one embodiment;
FIG. 9 is an internal block diagram of a computer device as a terminal in one embodiment;
Fig. 10 is an internal structural diagram of a computer device as a server in one embodiment.
Detailed Description
In order to make the objects, technical solutions and advantages of the present application more apparent, the present application will be further described in detail with reference to the accompanying drawings and examples. It should be understood that the specific embodiments described herein are for purposes of illustration only and are not intended to limit the present application.
The loss prediction method of the power system provided by the embodiment of the application can be applied to an application environment shown in fig. 1. Wherein the terminal 102 communicates with the server 104 via a network. The data storage system may store data that the server 104 needs to process. The data storage system may be integrated on the server 104 or may be located on a cloud or other network server. The terminal 102 acquires predicted power data and actual power data of the power system; determining a target loss prediction model, correcting an output layer of an initial loss prediction model according to actual power data by the target loss prediction model, and training the initial loss prediction model according to predicted power data; and acquiring real-time load data, inputting the real-time load data into a target loss prediction model to predict the energy loss of the power system, and obtaining a real-time loss predicted value. The terminal 102 may be, but not limited to, various personal computers, notebook computers, smart phones, tablet computers, internet of things devices, and portable wearable devices, where the internet of things devices may be smart speakers, smart televisions, smart air conditioners, smart vehicle devices, and the like. The portable wearable device may be a smart watch, smart bracelet, headset, or the like. The server 104 may be implemented as a stand-alone server or as a server cluster of multiple servers.
In an exemplary embodiment, as shown in fig. 2, a method for predicting loss of a power system is provided, and the method is applied to the terminal 102 in fig. 1, for example, and includes the following steps 202 to 206. Wherein:
step 202, obtaining predicted power data and actual power data of a power system.
The predicted power data is data obtained by predicting a load value and a loss value of the power system according to characteristic data such as meteorological data in a historical period. The actual power data is data obtained by acquiring the load value and the loss value of the power system in the actual operation process of the power system in the historical time period.
The load prediction model is determined, a load prediction data set obtained by predicting the load of each node in the power system in a historical time period by the load prediction model is read, wherein the load of each node refers to the electric energy required by the node at a certain moment, a generator set is a device for converting electric energy in the power system, and the output power of the generator set refers to the generating capacity of the generator set. In the economic dispatching process of the power system, the output power of each generator set is determined when the total cost of the power system is minimum based on each load predicted value in the load predicted data set, and the generator set output power data set is obtained. And (3) calling an alternating current power flow calculation method, and calculating a loss value of the power system corresponding to the generator set output power data set corresponding to each load predicted value to obtain a loss predicted data set.
The method comprises the steps of reading the actual load quantity of each node of the power system in a historical time period as an actual load data set, reading the actual output power data set of each generator set, calling an alternating current power flow calculation method, calculating energy loss corresponding to elements such as resistance, inductance and capacitance in the power system according to the actual load data set and the actual output power data set, and further determining the total actual loss value of the power system to obtain an actual loss data set. The actual load data set and the actual loss data set are taken as actual power data of the power system.
And 204, determining a target loss prediction model, wherein the target loss prediction model is obtained by correcting an output layer of an initial loss prediction model according to actual power data, and the initial loss prediction model is obtained by training according to predicted power data.
The initial loss prediction model comprises an input layer, a hidden layer, an output layer and an activation function layer, wherein the input layer receives original data or characteristic data as the input of the model; the hidden layers are connected with the output of the input layer and the input of the output layer, each hidden layer comprises one or more layers, each hidden layer comprises a plurality of neurons or a plurality of nodes, and the nodes are connected with each other through weights; the output layer outputs a final loss prediction result according to the output of the hidden layer; the activation function layer maps the input data to a nonlinear output data range by setting a nonlinear activation function. Model correction is used to achieve improved consistency of model output with actual operating data by adjusting the model output.
Illustratively, as shown in fig. 3, the initial loss prediction model is a regression model, specifically, an input layer of the initial loss prediction model acquires a load prediction data set, a hidden layer performs feature extraction on the load prediction data set, and an activation function layer transforms a weighted sum of the input data by using a linear activation function to obtain output data of an output layer. In the training process of the initial loss prediction model, a batch gradient descent method is adopted to carry out model training on the load prediction data set and the actual loss data set so as to achieve that the loss function value between output data obtained by the initial loss prediction model according to the input load prediction data set and the actual loss data set is minimum, and the initial loss prediction model is obtained. The input to output relationship of the initial loss prediction model is expressed as follows:
Y i ＝f i (w i Y i-1 +b i )
wherein Y is i-1 The output of the model i-1 layer is also the input of the model i layer, Y i For the output of the model i layer, w i Weights for model layer iVector, b i Is the bias vector of the ith layer of the model, f i Is an activation function of the model i layer.
The relation corresponding to the batch gradient descent method of the initial loss prediction model in the model training process is expressed as follows:
wherein h is i Is equal to w i And b i The function of the correlation is that,for the updated neuron output vector, α is the gradient descent step size.
Because the load prediction data set has errors with the actual load data set in the actual running process of the power system, the output data of the initial loss prediction model has errors with the actual loss data set in the actual running process of the power system, the output data of the initial loss prediction model can be adjusted according to the actual power data, and the data input layers and the data characteristic extraction layers of the previous layers of the model are fixed to realize the correction of the initial loss prediction model, and the target loss prediction model is obtained to improve the accuracy of the loss prediction result.
And 206, acquiring real-time load data, and inputting the real-time load data into a target loss prediction model to predict the energy loss of the power system, so as to obtain a real-time loss predicted value.
The method includes the steps of obtaining real-time load data output by a load prediction model, inputting the real-time load data into a target loss prediction model, and predicting energy loss of each node in the power system to obtain a real-time loss prediction value.
In the loss prediction method of the power system, the predicted power data and the actual power data of the power system are obtained; determining a target loss prediction model, correcting an output layer of an initial loss prediction model according to actual power data by the target loss prediction model, and training the initial loss prediction model according to predicted power data; and acquiring real-time load data, inputting the real-time load data into a target loss prediction model to predict the energy loss of the power system, and obtaining a real-time loss predicted value. According to the method, the output layer parameters of the initial loss model are corrected by adopting actual power data, so that the adaptation capacity and generalization capacity of the model can be improved, the training data of the initial loss prediction model are power prediction data, the problems of data noise, abnormal values or error marks and the like possibly exist, and the problems of data deviation, data errors, data integrity and the like can be avoided by correcting the initial loss prediction model by using the actual power data, so that the accuracy and the reliability of the prediction model are improved.
In one exemplary embodiment, wherein the predicted power data comprises a load prediction data set and a loss prediction data set; the actual power data includes an actual load data set and an actual loss data set; the target loss prediction model is obtained by correcting an output layer of the initial loss prediction model according to actual power data, and comprises the following steps: for each actual load value of the actual load data set, calculating the distance between the actual load value and each load predicted value of the load predicted data set, and taking the load predicted value with the smallest distance as a target load predicted value; determining an actual loss value corresponding to the actual load value in the actual loss data set; and correcting the output of the target load predicted value in the initial loss predicted model according to the actual loss value to obtain a target loss predicted model.
The distance refers to a euclidean distance, i.e. a straight line distance between two data points, and can be used to measure the similarity between two vectors in a vector space.
Illustratively, as described in the foregoing embodiment, the load prediction model predicts the load value over the historical period of time to obtain load prediction data, wherein the load prediction data includes the prediction time information. The predicted time information may be a load predicted value obtained by predicting a load value at an arbitrary time by the load prediction model. And determining an actual load value corresponding to the load predicted value in the actual load data set according to the predicted time information, and correcting the initial loss prediction model according to the actual load value and an actual loss value corresponding to the actual load value under the condition that the error between the load predicted value and the actual load value is larger than a preset threshold value. Specifically, the euclidean distance between the actual load value and each load predicted value in the load prediction data set is calculated to obtain a plurality of euclidean distance values, as shown in fig. 4, the load predicted value with the minimum euclidean distance is taken as a target load predicted value, the output value of the target load predicted value in the initial loss prediction model is obtained, the output value is corrected according to the actual loss value corresponding to the actual load value, and after model correction of each actual load value in the actual load data set is completed, the target loss prediction model is obtained.
In this embodiment, the euclidean distance between the actual value and the model predicted value is calculated, so that the degree of difference between the actual value and the predicted value can be obtained, and according to the degree of difference, the output of the model can be appropriately corrected, so that the deviation of the output is reduced, and the power loss value is further accurately obtained.
In one embodiment, the initial loss prediction model is trained from predicted power data, comprising: under each load predicted value, iteratively updating the output power of each generator set in the power system to obtain set output power data; and determining an initial loss prediction model, wherein the initial loss prediction model is obtained by training according to the predicted power data and the unit output power data.
The unit output power data comprise the output power of each generator unit in each iteration round number.
For example, as shown in fig. 5, the differential evolution algorithm may be used to iteratively update the planning scheme for the output power of each generator set in the power system according to the load prediction data to determine an optimal planning scheme for the output power of each generator set in the power system, where the optimal planning scheme is used to minimize the cost of the power system. And setting the crossover factor and the variation factor in the differential evolution algorithm as preset parameters with larger values because the power loss values corresponding to different power generation unit output power planning schemes are different, and reading the planning schemes of the output power of each power generation unit in the iterative updating process to obtain unit output power data. The power loss value is predicted according to the load prediction data and the unit output power data, so that the training data range of the initial loss prediction model can be expanded, and the initial loss prediction model applicable to different unit output power planning schemes is obtained.
In the embodiment, the output data of each generator set is read based on the differential evolution method, so that the data volume of model training data can be increased, the range and the quantity of input data can be expanded, the risk of excessive fitting is reduced, and the generalization capability and the prediction accuracy of the model are improved. The improvement of the data volume of the model training data is also beneficial to the model to learn different data characteristics better, so that the stability and the adaptability of the model are improved.
In one embodiment, for each actual load value of the actual load data set, calculating a distance between the actual load value and each load prediction value of the load prediction data set, and taking the load prediction value at which the distance is the smallest as the target load prediction value, includes: performing cluster analysis on the actual load data set to obtain a plurality of target actual load values; for each target actual load value, the distance between the target actual load value and each load predicted value is calculated, and the load predicted value at which the distance is the smallest is used as the target load predicted value.
The clustering analysis is a data analysis method, wherein data are divided into a plurality of groups of similar data according to a preset rule, so that the similarity among the data in the groups is higher than the similarity among the data in different groups.
Illustratively, as shown in fig. 6, in the actual load data set, a k-means clustering method is used in combination with a contour coefficient method to perform cluster analysis on the actual load data set, and the contour coefficient method is used for determining the target cluster category number of the actual load data set by calculating a contour coefficient, wherein the contour coefficient is used for representing compactness in a data set and separation degree among data sets. The k-means clustering method divides the actual load data set according to the number of target clustering categories, and obtains a target actual load value according to a preset rule that the similarity between data in the data sets is higher than the similarity between different data sets.
For each target actual load value, the Euclidean distance between the target actual load value and each load predicted value in the load predicted data set is calculated, and the load predicted value with the smallest distance is taken as the target load predicted value corresponding to the target actual load value. Because the target load predicted value adopts a differential evolution algorithm to amplify the output power data of the generator set in the training process of the initial loss predicted model, the differential evolution algorithm can also be adopted to amplify the data of the target actual load value, so that the output power of the target generator set corresponding to the target actual load value is obtained. And (3) invoking an alternating current power flow calculation method, calculating a power system loss value according to the target actual load value and the target generator set output power to obtain an actual loss data set, inputting the target actual load value, the target generator set output power and the actual loss data set into an initial loss prediction model, and correcting the output of the target load prediction value in the initial loss prediction model based on the actual loss value corresponding to the target actual load value.
In this embodiment, by performing cluster analysis on the actual load data set, the data may be divided into different data subsets, each data subset may be used to represent a specific class of load behavior or specific load mode, and based on the clustering result of the actual load data set, each data subset may be corrected pertinently in the model output correction process, so as to reduce the deviation of the prediction result. When different data subsets correspond to different behavior modes, the clustering analysis can improve the accuracy of model prediction, and change trend of actual load values can be reflected better and accurately, so that more reliable and accurate prediction results are provided.
In one embodiment, performing a cluster analysis on the actual load dataset to obtain a plurality of target actual load values, including: acquiring the number of clusters, and determining initial load cluster centers of a plurality of actual load data sets according to the number of clusters; calculating a second distance between the actual load value and each initial load clustering center according to each actual load value, and dividing the actual load values of the actual load data sets according to the second distances to obtain a plurality of actual load data subsets; and carrying out iterative updating on the initial load clustering center according to the average value of the load actual values in each actual load data subset, and obtaining each target actual load value under the condition that the preset termination condition is met.
The initial load clustering center is used for initializing an initial position or an initial value of the actual load data set clustering center when cluster analysis is performed. The clustering number refers to the preset number of actual load data subsets obtained by the clustering analysis when the clustering analysis is performed.
Specifically, according to the clustering quantity, the clustering center points are initialized randomly to obtain initial load clustering centers, the Euclidean distance from each actual load value in the actual load data set to the initial load clustering center is calculated for each initial load clustering center, the Euclidean distance from each actual load value to each initial load clustering center is obtained, the initial load clustering center with the smallest Euclidean distance corresponding to each actual load value is used as the initial load clustering center corresponding to the actual load value, and each actual load value corresponding to the initial load clustering center is used as the actual load data subset corresponding to the initial load clustering center. Calculating an average value of each actual load value in each actual load data subset, iteratively updating an initial load cluster center according to the average value of each actual load data subset, and under the condition that the average value in each actual load data subset is not changed any more, obtaining a target load cluster center corresponding to each actual load data subset. Because different clustering numbers correspond to different target load clustering centers, the contour coefficients of the different clustering numbers are calculated, and the contour coefficients are used for representing the compactness in the data sets and the separation degree among the data sets, the clustering number corresponding to the maximum contour coefficient is used as the target clustering number, the target load clustering centers corresponding to the target clustering number are output, and the target load clustering centers corresponding to the target clustering number are used as target actual load values, so that the clustering analysis of the actual load data set is realized.
In the embodiment, the clustering quantity can be automatically determined through a k-means clustering algorithm and a contour coefficient method, and the distance between each data set is optimized, so that the accuracy of a clustering result is improved, too discrete clustering or too fuzzy classification is avoided, and the clustering resolution is further improved.
In one embodiment, calculating a second distance of the actual load value from each initial load cluster center includes: calculating the difference value between the actual load value and each initial load clustering center to obtain a plurality of difference values corresponding to each actual load value; carrying out average value operation on the square value of each difference value to obtain a mean square value; and performing square root operation on the mean square value to obtain a second distance.
In the process of carrying out clustering analysis on the actual load data set, calculating the difference value between each actual load value in the actual load data set and the initial load clustering center, calculating the mean value of the square value of each difference value, and finally carrying out square root operation on the mean value to obtain a second distance. Wherein, the calculation formula of the second distance is expressed as follows:
wherein d (x, y) is Euclidean distance, x i For the actual load value, y i For the initial load cluster center, n is the number of data samples.
In this embodiment, the euclidean distance is used to represent the degree of difference between the data points, the mean variance is used as the feature statistic of the data, so that the information of the data distribution in the sample can be provided, the covariance relation between the features is weighted and considered, the square root of the mean square difference is calculated, the calculation mode is simple, and the relative similarity between the data points is reflected more accurately.
In one exemplary embodiment, as shown in fig. 7, a method of loss prediction for an electrical power system is provided, the method comprising steps 702 to 716. Wherein:
step 702, obtaining predicted power data and actual power data of a power system; wherein the predicted power data includes a load prediction data set and a loss prediction data set; the actual power data includes an actual load data set and an actual loss data set.
And step 704, under each load predicted value, iteratively updating the output power of each generator set in the power system to obtain set output power data.
Step 706, determining an initial loss prediction model, wherein the initial loss prediction model is obtained through training according to the predicted power data and the unit output power data.
Step 708, obtaining the number of clusters, and determining initial load cluster centers of a plurality of actual load data sets according to the number of clusters; calculating the difference value between the actual load value and each initial load clustering center to obtain a plurality of difference values corresponding to each actual load value; carrying out average value operation on the square value of each difference value to obtain a mean square value; and performing square root operation on the mean square value to obtain a second distance.
Step 710, dividing the actual load values of the actual load data set according to the second distance to obtain a plurality of actual load data subsets; and carrying out iterative updating on the initial load clustering center according to the average value of the load actual values in each actual load data subset, and obtaining each target actual load value under the condition that the preset termination condition is met.
Step 712, calculating a distance between the target actual load value and each load predicted value for each target actual load value, and setting the load predicted value at the minimum distance as the target load predicted value.
Step 714, determining an actual loss value corresponding to the actual load value in the actual loss data set; and correcting the output of the target load predicted value in the initial loss predicted model according to the actual loss value to obtain a target loss predicted model.
Step 716, acquiring real-time load data, and inputting the real-time load data into a target loss prediction model to predict the energy loss of the power system, thereby obtaining a real-time loss predicted value.
In the embodiment, the predicted power data and the actual power data of the power system are obtained; determining a target loss prediction model, correcting an output layer of an initial loss prediction model according to actual power data by the target loss prediction model, and training the initial loss prediction model according to predicted power data; and acquiring real-time load data, inputting the real-time load data into a target loss prediction model to predict the energy loss of the power system, and obtaining a real-time loss predicted value. According to the method, the output layer parameters of the initial loss model are corrected by adopting actual power data, so that the adaptation capacity and generalization capacity of the model can be improved, the training data of the initial loss prediction model are power prediction data, the problems of data noise, abnormal values or error marks and the like possibly exist, and the problems of data deviation, data errors, data integrity and the like can be avoided by correcting the initial loss prediction model by using the actual power data, so that the accuracy and the reliability of the prediction model are improved.
It should be understood that, although the steps in the flowcharts related to the embodiments described above are sequentially shown as indicated by arrows, these steps are not necessarily sequentially performed in the order indicated by the arrows. The steps are not strictly limited to the order of execution unless explicitly recited herein, and the steps may be executed in other orders. Moreover, at least some of the steps in the flowcharts described in the above embodiments may include a plurality of steps or a plurality of stages, which are not necessarily performed at the same time, but may be performed at different times, and the order of the steps or stages is not necessarily performed sequentially, but may be performed alternately or alternately with at least some of the other steps or stages.
Based on the same inventive concept, the embodiment of the application also provides a loss prediction device of the electric power system for realizing the loss prediction method of the electric power system. The implementation of the solution provided by the device is similar to the implementation described in the above method, so the specific limitation in the embodiments of the loss prediction device for one or more power systems provided below may refer to the limitation of the loss prediction method for a power system, which is not described herein.
In one exemplary embodiment, as shown in the loss prediction of a power system, there is provided a loss prediction apparatus 800 of a power system, including: a data acquisition module 802, a model acquisition module 804, and a loss prediction module 806, wherein:
a data acquisition module 802, configured to acquire predicted power data and actual power data of the power system;
the model obtaining module 804 is configured to determine a target loss prediction model, where the target loss prediction model is obtained by correcting an output layer of an initial loss prediction model according to actual power data, and the initial loss prediction model is obtained by training according to predicted power data;
the loss prediction module 806 is configured to obtain real-time load data, input the real-time load data into a target loss prediction model to predict energy loss of the power system, and obtain a real-time loss prediction value.
In one embodiment, wherein the predicted power data comprises a load prediction data set and a loss prediction data set; the actual power data includes an actual load data set and an actual loss data set; the model obtaining module 804 is further configured to calculate, for each actual load value of the actual load data set, a distance between the actual load value and each load prediction value of the load prediction data set, and use the load prediction value when the distance is minimum as a target load prediction value; determining an actual loss value corresponding to the actual load value in the actual loss data set; and correcting the output of the target load predicted value in the initial loss predicted model according to the actual loss value to obtain a target loss predicted model.
In one embodiment, the model obtaining module 804 is further configured to iteratively update, under each load predicted value, output power of each generator set in the power system to obtain set output power data; and determining an initial loss prediction model, wherein the initial loss prediction model is obtained by training according to the predicted power data and the unit output power data.
In one embodiment, the model obtaining module 804 is further configured to perform cluster analysis on the actual load data set to obtain a plurality of target actual load values; for each target actual load value, the distance between the target actual load value and each load predicted value is calculated, and the load predicted value at which the distance is the smallest is used as the target load predicted value.
In one embodiment, the model obtaining module 804 is further configured to obtain a number of clusters, and determine initial load cluster centers of the plurality of actual load data sets according to the number of clusters; calculating a second distance between the actual load value and each initial load clustering center according to each actual load value, and dividing the actual load values of the actual load data sets according to the second distances to obtain a plurality of actual load data subsets; and carrying out iterative updating on the initial load clustering center according to the average value of the load actual values in each actual load data subset, and obtaining each target actual load value under the condition that the preset termination condition is met.
In one embodiment, the model obtaining module 804 is further configured to calculate a difference between the actual load value and each initial load cluster center, so as to obtain a plurality of differences corresponding to each actual load value; carrying out average value operation on the square value of each difference value to obtain a mean square value; and performing square root operation on the mean square value to obtain a second distance.
The respective modules in the loss prediction apparatus of the electric power system described above may be implemented in whole or in part by software, hardware, and combinations thereof. The above modules may be embedded in hardware or may be independent of a processor in the computer device, or may be stored in software in a memory in the computer device, so that the processor may call and execute operations corresponding to the above modules.
In one exemplary embodiment, a computer device is provided, which may be a server, the internal structure of which may be as shown in fig. 9. The computer device includes a processor, a memory, an Input/Output interface (I/O) and a communication interface. The processor, the memory and the input/output interface are connected through a system bus, and the communication interface is connected to the system bus through the input/output interface. Wherein the processor of the computer device is configured to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, computer programs, and a database. The internal memory provides an environment for the operation of the operating system and computer programs in the non-volatile storage media. The database of the computer device is used to store loss prediction data for the power system. The input/output interface of the computer device is used to exchange information between the processor and the external device. The communication interface of the computer device is used for communicating with an external terminal through a network connection. The computer program, when executed by a processor, implements a method of loss prediction for an electrical power system.
In an exemplary embodiment, a computer device, which may be a terminal, is provided, and an internal structure thereof may be as shown in fig. 10. The computer device includes a processor, a memory, an input/output interface, a communication interface, a display unit, and an input means. The processor, the memory and the input/output interface are connected through a system bus, and the communication interface, the display unit and the input device are connected to the system bus through the input/output interface. Wherein the processor of the computer device is configured to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of the operating system and computer programs in the non-volatile storage media. The input/output interface of the computer device is used to exchange information between the processor and the external device. The communication interface of the computer device is used for carrying out wired or wireless communication with an external terminal, and the wireless mode can be realized through WIFI, a mobile cellular network, NFC (near field communication) or other technologies. The computer program, when executed by a processor, implements a method of loss prediction for an electrical power system. The display unit of the computer device is used for forming a visual picture, and can be a display screen, a projection device or a virtual reality imaging device. The display screen can be a liquid crystal display screen or an electronic ink display screen, and the input device of the computer equipment can be a touch layer covered on the display screen, can also be a key, a track ball or a touch pad arranged on the shell of the computer equipment, and can also be an external keyboard, a touch pad or a mouse and the like.
It will be appreciated by persons skilled in the art that the foregoing structures, which are merely block diagrams of partial structures related to the aspects of the present application, are not limiting of the computer device to which the aspects of the present application may be applied, and that a particular computer device may include more or fewer components than shown, or may combine certain components, or have different arrangements of components.
In one embodiment, a computer device is provided, comprising a memory and a processor, the memory having stored therein a computer program, the processor implementing the steps of the method embodiments described above when the computer program is executed.
In one embodiment, a computer-readable storage medium is provided, on which a computer program is stored which, when executed by a processor, implements the steps of the method embodiments described above.
In an embodiment, a computer program product is provided, comprising a computer program which, when executed by a processor, implements the steps of the method embodiments described above.
It should be noted that, the user information (including, but not limited to, user equipment information, user personal information, etc.) and the data (including, but not limited to, data for analysis, stored data, presented data, etc.) referred to in the present application are information and data authorized by the user or sufficiently authorized by each party, and the collection, use, and processing of the related data are required to meet the related regulations.
Those skilled in the art will appreciate that implementing all or part of the above described methods may be accomplished by way of a computer program stored on a non-transitory computer readable storage medium, which when executed, may comprise the steps of the embodiments of the methods described above. Any reference to memory, database, or other medium used in the various embodiments provided herein may include at least one of non-volatile and volatile memory. The nonvolatile Memory may include Read-Only Memory (ROM), magnetic tape, floppy disk, flash Memory, optical Memory, high density embedded nonvolatile Memory, resistive random access Memory (ReRAM), magnetic random access Memory (Magnetoresistive Random Access Memory, MRAM), ferroelectric Memory (Ferroelectric Random Access Memory, FRAM), phase change Memory (Phase Change Memory, PCM), graphene Memory, and the like. Volatile memory can include random access memory (Random Access Memory, RAM) or external cache memory, and the like. By way of illustration, and not limitation, RAM can be in the form of a variety of forms, such as static random access memory (Static Random Access Memory, SRAM) or dynamic random access memory (Dynamic Random Access Memory, DRAM), and the like. The databases referred to in the various embodiments provided herein may include at least one of relational databases and non-relational databases. The non-relational database may include, but is not limited to, a blockchain-based distributed database, and the like. The processors referred to in the embodiments provided herein may be general purpose processors, central processing units, graphics processors, digital signal processors, programmable logic units, quantum computing-based data processing logic units, etc., without being limited thereto.
The technical features of the above embodiments may be arbitrarily combined, and all possible combinations of the technical features in the above embodiments are not described for brevity of description, however, as long as there is no contradiction between the combinations of the technical features, they should be considered as the scope of the description.
The above examples only represent a few embodiments of the present application, which are described in more detail and are not to be construed as limiting the scope of the present application. It should be noted that it would be apparent to those skilled in the art that various modifications and improvements could be made without departing from the spirit of the present application, which would be within the scope of the present application. Accordingly, the scope of protection of the present application shall be subject to the appended claims.
Claims (10)
1. A method of loss prediction for an electrical power system, the method comprising:
acquiring predicted power data and actual power data of a power system;
determining a target loss prediction model, wherein the target loss prediction model is obtained by correcting an output layer of an initial loss prediction model according to the actual power data, and the initial loss prediction model is obtained by training the predicted power data;
And acquiring real-time load data, inputting the real-time load data into the target loss prediction model to predict the energy loss of the power system, and obtaining a real-time loss predicted value.
2. The method of claim 1, wherein the predicted power data comprises a load prediction data set and a loss prediction data set; the actual power data comprises an actual load data set and an actual loss data set; the target loss prediction model is obtained by correcting an output layer of an initial loss prediction model according to the actual power data, and comprises the following steps:
for each actual load value of the actual load data set, calculating the distance between the actual load value and each load predicted value of the load predicted data set, and taking the load predicted value with the smallest distance as a target load predicted value;
determining an actual loss value corresponding to the actual load value in the actual loss data set;
and correcting the output of the target load predicted value at the initial loss predicted model according to the actual loss value to obtain the target loss predicted model.
3. The method of claim 2, wherein the initial loss prediction model is trained from the predicted power data, comprising:
Under each load predicted value, iteratively updating the output power of each generator set in the power system to obtain set output power data;
and determining the initial loss prediction model, wherein the initial loss prediction model is obtained through training according to the predicted power data and the unit output power data.
4. The method according to claim 2, wherein the calculating, for each actual load value of the actual load data set, a distance between the actual load value and each load prediction value of the load prediction data set, and taking the load prediction value at which the distance is the smallest as the target load prediction value, includes:
performing cluster analysis on the actual load data set to obtain a plurality of target actual load values;
and calculating the distance between the target actual load value and each load predicted value for each target actual load value, and taking the load predicted value with the smallest distance as the target load predicted value.
5. The method of claim 4, wherein performing cluster analysis on the actual load dataset to obtain a plurality of target actual load values comprises:
acquiring the number of clusters, and determining initial load cluster centers of a plurality of actual load data sets according to the number of clusters;
Calculating a second distance between the actual load value and each initial load clustering center for each actual load value, and dividing the actual load value of the actual load data set according to the second distance to obtain a plurality of actual load data subsets;
and carrying out iterative updating on the initial load clustering center according to the average value of the load actual values in each actual load data subset, and obtaining each target actual load value under the condition that the preset termination condition is met.
6. The method of claim 4, wherein said calculating a second distance of said actual load value from each of said initial load cluster centers comprises:
calculating the difference value between the actual load value and each initial load clustering center to obtain a plurality of difference values corresponding to each actual load value;
carrying out average value operation on the square value of each difference value to obtain a mean square value;
and carrying out square root operation on the mean square value to obtain the second distance.
7. A loss prediction apparatus for an electric power system, the apparatus comprising:
the data acquisition module is used for acquiring predicted power data and actual power data of the power system;
The model acquisition module is used for determining a target loss prediction model, the target loss prediction model is obtained by correcting an output layer of an initial loss prediction model according to the actual power data, and the initial loss prediction model is obtained by training according to the predicted power data;
and the loss prediction module is used for acquiring real-time load data, inputting the real-time load data into the target loss prediction model to predict the energy loss of the power system, and obtaining a real-time loss predicted value.
8. A computer device comprising a memory and a processor, the memory storing a computer program, characterized in that the processor implements the steps of the method of any of claims 1 to 6 when the computer program is executed.
9. A computer readable storage medium, on which a computer program is stored, characterized in that the computer program, when being executed by a processor, implements the steps of the method of any of claims 1 to 6.
10. A computer program product comprising a computer program, characterized in that the computer program, when being executed by a processor, implements the steps of the method of any of claims 1 to 6.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311565971.6A CN117638882A (en) | 2023-11-22 | 2023-11-22 | Loss prediction method, device, equipment, medium and product of electric power system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311565971.6A CN117638882A (en) | 2023-11-22 | 2023-11-22 | Loss prediction method, device, equipment, medium and product of electric power system |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN117638882A true CN117638882A (en) | 2024-03-01 |
Family
ID=90019195
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202311565971.6A Pending CN117638882A (en) | 2023-11-22 | 2023-11-22 | Loss prediction method, device, equipment, medium and product of electric power system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117638882A (en) |
-
2023
- 2023-11-22 CN CN202311565971.6A patent/CN117638882A/en active Pending
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| WO2022105108A1 (en) | Network data classification method, apparatus, and device, and readable storage medium | |
| CN113887845B (en) | Extreme event prediction method, device, equipment and storage medium | |
| CN112131261A (en) | Community query method and device based on community network and computer equipment | |
| CN117371510A (en) | Power load prediction model training method and device based on domain generalization | |
| WO2024082767A1 (en) | Gain adjustment method and apparatus, computer device and storage medium | |
| CN115392594B (en) | Electrical load model training method based on neural network and feature screening | |
| CN116308644A (en) | Recommendation method, device, equipment, storage medium and product of target product | |
| CN117638882A (en) | Loss prediction method, device, equipment, medium and product of electric power system | |
| CN116191398A (en) | Load prediction method, load prediction device, computer equipment and storage medium | |
| CN114723147A (en) | New energy power prediction method based on improved wavelet transform and neural network | |
| US20220121926A1 (en) | Tensor ring decomposition for neural networks | |
| CN116821720A (en) | User clustering method, device, computer equipment and storage medium | |
| CN117312892A (en) | User clustering method, device, computer equipment and storage medium | |
| CN116127183A (en) | Service recommendation method, device, computer equipment and storage medium | |
| CN117574151A (en) | Target recognition model training method and device, storage medium and computer equipment | |
| CN116628538A (en) | Patient clustering method and device based on graph alignment neural network and computer equipment | |
| CN117633425A (en) | Power equipment image processing method, device, equipment and medium | |
| CN115880023A (en) | Size recommendation method and device, computer equipment and storage medium | |
| CN115564560A (en) | Risk assessment method, apparatus and storage medium | |
| CN117909714A (en) | Charging power prediction model training method and device and computer equipment | |
| CN117017277A (en) | Identity recognition method, device, equipment and medium based on photoelectric volume pulse wave | |
| CN117829892A (en) | Three-dimensional model supply and demand analysis method, device, computer equipment and storage medium | |
| CN116029447A (en) | Predictive model training method and device, medium and computer equipment | |
| CN116933918A (en) | Electric power data prediction method for internet financial service | |
| CN117407642A (en) | Digital transformation data processing method, device, equipment and storage medium |
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 |