CN116108974A - Demand response baseline load prediction method and device considering meteorological factors - Google Patents
Demand response baseline load prediction method and device considering meteorological factors Download PDFInfo
- Publication number
- CN116108974A CN116108974A CN202211683644.6A CN202211683644A CN116108974A CN 116108974 A CN116108974 A CN 116108974A CN 202211683644 A CN202211683644 A CN 202211683644A CN 116108974 A CN116108974 A CN 116108974A
- Authority
- CN
- China
- Prior art keywords
- data
- load
- historical
- baseline
- prediction
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—Systems or methods specially adapted for specific business sectors, e.g. utilities or tourism
- G06Q50/06—Electricity, gas or water supply
-
- H—ELECTRICITY
- H02—GENERATION; CONVERSION OR DISTRIBUTION OF ELECTRIC POWER
- H02J—CIRCUIT ARRANGEMENTS OR SYSTEMS FOR SUPPLYING OR DISTRIBUTING ELECTRIC POWER; SYSTEMS FOR STORING ELECTRIC ENERGY
- H02J3/00—Circuit arrangements for ac mains or ac distribution networks
- H02J3/003—Load forecast, e.g. methods or systems for forecasting future load demand
-
- 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
Abstract
The invention discloses a demand response baseline load prediction method and device considering meteorological factors, wherein the method comprises the following steps: acquiring historical load data; determining abnormal data in the historical load data based on the tree forest constructed by the historical load data; filling the missing data after deleting the abnormal data in the historical load data based on a KNN algorithm; determining a primary one of the weather factors associated with the baseline load prediction based on the principal component analysis; and carrying out baseline load prediction based on a model constructed by adopting filled historical load data and main meteorological factor data. The identification of abnormal data and the filling of missing data are realized through the constructed tree forest and KNN algorithm; meanwhile, the influence of meteorological factors on load prediction is considered, so that the finally obtained baseline load prediction result is more accurate, and the factors with larger contribution rate in the meteorological factors are analyzed by using a principal component analysis method, so that the defect of low algorithm calculation speed caused by considering all meteorological factors is avoided.
Description
Technical Field
The invention relates to the technical field of load prediction, in particular to a demand response baseline load prediction method and device considering meteorological factors.
Background
The user baseline load is an important reference for the user to participate in the execution effect of the demand response project, and is influenced by various factors such as load environment, user electricity behavior habit and the like. The baseline load provides a data reference for quantitatively evaluating the load reduction degree of the user, is the basis for the demand response implementation mechanism to carry out rewards and punishments on clients, and is a measurement index for evaluating the implementation effect of the demand response project. However, the demand response implementation mechanism can only obtain the load data of the user after the demand response, and cannot obtain the baseline load data of the user when the user does not respond to the demand, so that the specific demand response quantity of the user cannot be known.
Accurate baseline load estimation is important to the implementation of incentive type demand responses because it directly impacts the economic benefits of both the demand response implementer and the participants. In the excitation type demand response, it is necessary to estimate the baseline load at two different spatial levels: individuals and clusters. The individual-level baseline load refers to the baseline load of an individual user, and the estimation result is mainly used for compensation settlement between a load aggregator and a demand response participant; the cluster baseline load (aggregated baseline load, ABL) refers to the sum of all CBLs (customer baseline load) of the load aggregator agent, and the estimation result is not only the basis for settlement between the system operator and the load aggregator, but also the basis for quantifying the implementation effect of the whole demand response project.
The traditional demand response baseline load prediction method generally obtains a predicted data basis by processing sample load data, for example, a typical scale factor correction method is adopted to select the sample load data, and a load prediction method based on a combination of time sequence and Kalman filtering is adopted, so that factors considered in prediction are not comprehensive enough. In the prior art, when the demand response baseline load is predicted, the influence of meteorological factors on the adjustable load in the load is not considered, and the baseline load is predicted only through historical load data, so that the precision of the finally obtained baseline load is not high enough.
Disclosure of Invention
In view of the above, the embodiment of the invention provides a method and a device for predicting a demand response baseline load in consideration of meteorological factors, so as to solve the technical problem of low baseline load precision in the prior art.
The technical scheme provided by the invention is as follows:
an embodiment of the present invention provides a method for predicting a demand response baseline load in consideration of meteorological factors, including: acquiring historical load data; determining abnormal data in the historical load data based on a tree forest constructed by the historical load data; filling the missing data after deleting the abnormal data in the historical load data based on a KNN algorithm; determining a primary one of the weather factors associated with the baseline load prediction based on the principal component analysis; and carrying out baseline load prediction based on a model constructed by adopting filled historical load data and main meteorological factor data.
Optionally, determining the abnormal data in the historical load data based on the tree forest constructed by the historical load data includes: determining root nodes and leaf nodes of a tree forest based on the relation between the historical load data and preset cut point values; calculating the height of each tree in the tree forest of each historical load data; calculating an anomaly score for each historical load data based on the average value of all heights and the preset cutting point data; determining whether the historical load data is abnormal data based on an abnormal score.
Optionally, determining the root node and the leaf node of the tree forest based on the historical load data and the relation between the historical load data and the preset cut point value comprises: taking sample data in the historical load data as a root node of each tree in the tree forest; placing any dimension data in a left leaf node or a right leaf node based on the size relation between any dimension data in the sample data and a preset cutting point value; and stopping generating new leaf nodes when each tree in the tree forest meets the preset condition.
Optionally, filling the missing data after deleting the abnormal data in the historical load data based on a KNN algorithm includes: deleting the determined abnormal data to generate missing data; calculating the Euclidean distance of each missing data based on a data set matrix constructed by the historical load data; and calculating the substitution value of the missing data based on the k neighbor data selected by the Euclidean distance and the weight of the k neighbor data to fill.
Optionally, determining a major one of the weather factors associated with the baseline load prediction based on the principal component analysis includes: carrying out standardization processing on meteorological factor data related to baseline load prediction; calculating the characteristic value and the characteristic vector of the weather factor data after standardized processing; the primary weather factor is determined based on the principal component scores calculated using the eigenvectors and the cumulative contribution calculated using the eigenvalues.
Optionally, the baseline load prediction based on a model constructed using the padded historical load data and the primary meteorological factor data includes: decoupling the filled historical load data to obtain non-adjustable load data, and adjustable load data which does not contain air conditioner load; and carrying out baseline load prediction by adopting a model constructed by non-adjustable load data, adjustable load data without air conditioner load, adjustable load data and main meteorological factor data.
Optionally, the baseline load prediction is performed using a model constructed from non-adjustable load data, adjustable load data that does not include air conditioning load, adjustable load data, and primary meteorological factor data, including: the method comprises the steps of constructing a proportion prediction model by taking total load data in historical loads, adjustable load data which does not contain air conditioning loads and main meteorological factor data as input and taking total load proportion occupied by the sum of the non-adjustable loads and the air conditioning loads as output; constructing a load prediction model by taking non-adjustable load, air conditioner load and main meteorological factor data as input and taking a baseline load value of the non-adjustable load and the air conditioner load as output; and determining the total baseline load based on the ratio of the baseline load predicted by the load prediction model and the proportion predicted by the proportion prediction model.
A second aspect of the embodiments of the present invention provides a device for predicting a demand response baseline load in consideration of weather factors, including: the data acquisition module is used for acquiring historical load data; the abnormal data judging module is used for determining abnormal data in the historical load data based on the tree forest constructed by the historical load data; the filling module is used for filling the missing data after deleting the abnormal data in the historical load data based on a KNN algorithm; a principal component analysis module for determining a primary one of the weather factors associated with the baseline load prediction based on the principal component analysis; and the prediction module is used for predicting the baseline load based on a model constructed by adopting the filled historical load data and the main meteorological factor data.
A third aspect of the embodiments of the present invention provides a computer readable storage medium storing computer instructions for causing a computer to execute the method for predicting a demand response baseline load taking into account meteorological factors according to any one of the first aspect and the first aspect of the embodiments of the present invention.
A fourth aspect of an embodiment of the present invention provides an electronic device, including: the system comprises a memory and a processor, wherein the memory and the processor are in communication connection, the memory stores computer instructions, and the processor executes the computer instructions so as to execute the weather factor considered demand response baseline load prediction method according to any one of the first aspect and the first aspect of the embodiment of the invention.
The technical scheme provided by the invention has the following effects:
according to the method and the device for predicting the demand response baseline load taking meteorological factors into consideration, historical load data are obtained; determining abnormal data in the historical load data based on the tree forest constructed by the historical load data; filling the missing data after deleting the abnormal data in the historical load data based on a KNN algorithm; determining a primary one of the weather factors associated with the baseline load prediction based on the principal component analysis; and carrying out baseline load prediction based on a model constructed by adopting filled historical load data and main meteorological factor data. Therefore, the recognition and filling of the abnormal data are realized through the constructed tree forest and KNN algorithm, so that the subsequent prediction result is more accurate; meanwhile, the influence of meteorological factors on load prediction is considered in the process of carrying out demand response baseline load prediction, so that the finally obtained baseline load prediction result is more accurate, and in addition, the factors with larger contribution rate in the meteorological factors affecting the load prediction are analyzed by using a principal component analysis method, so that the defect of slow algorithm calculation speed caused by considering all the meteorological factors is avoided.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings that are needed in the description of the embodiments or the prior art will be briefly described, and it is obvious that the drawings in the description below are some embodiments of the present invention, and other drawings can be obtained according to the drawings without inventive effort for a person skilled in the art.
FIG. 1 is a flow chart of a method of demand response baseline load prediction that considers meteorological factors in accordance with an embodiment of the present invention;
FIG. 2 is a schematic diagram of a contribution rate of a main component of a weather indicator according to an embodiment of the invention;
FIG. 3 is a flow chart of a method of demand response baseline load prediction that considers meteorological factors in accordance with another embodiment of the present invention;
FIG. 4 is a block diagram of a demand response baseline load prediction device that accounts for weather factors, according to an embodiment of the invention;
FIG. 5 is a schematic diagram of a computer-readable storage medium provided according to an embodiment of the present invention;
fig. 6 is a schematic structural diagram of an electronic device according to an embodiment of the present invention.
Detailed Description
In order that those skilled in the art will better understand the present invention, a technical solution in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in which it is apparent that the described embodiments are only some embodiments of the present invention, not all embodiments. All other embodiments, which can be made by those skilled in the art based on the embodiments of the present invention without making any inventive effort, shall fall within the scope of the present invention.
The terms first, second, third, fourth and the like in the description and in the claims and in the above drawings are used for distinguishing between similar objects and not necessarily for describing a particular sequential or chronological order. It is to be understood that the data so used may be interchanged where appropriate such that the embodiments described herein may be implemented in other sequences than those illustrated or otherwise described herein. Furthermore, the terms "comprises," "comprising," and "having," and any variations thereof, are intended to cover a non-exclusive inclusion, such that a process, method, system, article, or apparatus that comprises a list of steps or elements is not necessarily limited to those steps or elements expressly listed but may include other steps or elements not expressly listed or inherent to such process, method, article, or apparatus.
In accordance with an embodiment of the present invention, there is provided a method of demand response baseline load prediction taking into account meteorological factors, it being noted that the steps illustrated in the flow chart of the drawings may be performed in a computer system, such as a set of computer executable instructions, and, although a logical sequence is illustrated in the flow chart, in some cases, the steps illustrated or described may be performed in a different order than that illustrated herein.
In this embodiment, a method for predicting a demand response baseline load considering meteorological factors is provided, which may be used for electronic devices, such as computers, mobile phones, tablet computers, etc., fig. 1 is a flowchart of a method for predicting a demand response baseline load considering meteorological factors according to an embodiment of the present invention, as shown in fig. 1, and the method includes the following steps:
step S101: historical load data is obtained. Specifically, when the baseline load prediction needs to be performed on the commercial building, the historical load data may be the obtained historical load data of the commercial building user participating in the demand response, and when the data is obtained, the historical load data may be directly obtained from a load database of the corresponding commercial building user. For example, the obtained historical load data may include load data of a plurality of days, and as sample data, load data of a plurality of times are included in the load data of each day, wherein if one load data is obtained every 15 minutes, 96 load data are included in each sample data.
Step S102: and determining abnormal data in the historical load data based on the tree forest constructed by the historical load data. Specifically, constructing a tree forest by adopting load data at each moment contained in each sample data in the historical load data, namely adopting the load data at each moment as a node of each tree in the tree forest; and then determining whether the historical load data is abnormal data or not by judging the height of each load data on each tree.
Step S103: and filling the missing data after deleting the abnormal data in the historical load data based on a KNN algorithm. In particular, the KNN algorithm (K nearest neighbor) means that if a sample belongs to a certain class for the majority of K most similar (i.e., nearest neighbor) samples in the feature space, then the sample also belongs to that class. The method only determines the category to which the sample to be classified belongs according to the category of one or more samples which are nearest to each other in the classification decision. Therefore, k neighbor data of the abnormal data can be determined through the KNN algorithm, the k neighbor data is weighted, the substitution value of the missing data is determined through the weighted calculation, and the missing data is filled.
Step S104: the main meteorological factors of the meteorological factors related to the baseline load prediction are determined based on a principal component analysis method. Specifically, the weather factors related to the baseline load prediction include, in particular, barometric pressure, air temperature, humidity, wind speed, precipitation, and sunlight, and the main weather factors are determined by performing principal component analysis on these weather factors. That is, a weather index having a contribution to the baseline load of more than 80% can be selected as reference data for baseline load prediction from among weather factors by a principal component analysis method.
Step S105: and carrying out baseline load prediction based on a model constructed by adopting filled historical load data and main meteorological factor data. Specifically, the filled historical load data and main meteorological factor data can be used for training a neural network model such as a BP neural network model, parameters of the neural network are adjusted, and the obtained model can be used for predicting the baseline load.
According to the method for predicting the demand response baseline load taking meteorological factors into consideration, historical load data are obtained; determining abnormal data in the historical load data based on the tree forest constructed by the historical load data; filling the missing data after deleting the abnormal data in the historical load data based on a KNN algorithm; determining a primary one of the weather factors associated with the baseline load prediction based on the principal component analysis; and carrying out baseline load prediction based on a model constructed by adopting filled historical load data and main meteorological factor data. Therefore, the recognition and filling of the abnormal data are realized through the constructed tree forest and KNN algorithm, so that the subsequent prediction result is more accurate; meanwhile, the influence of meteorological factors on load prediction is considered in the process of carrying out demand response baseline load prediction, so that the finally obtained baseline load prediction result is more accurate, and in addition, the factors with larger contribution rate in the meteorological factors affecting the load prediction are analyzed by using a principal component analysis method, so that the defect of slow algorithm calculation speed caused by considering all the meteorological factors is avoided.
In one embodiment, determining abnormal data in the historical load data based on a tree forest constructed from the historical load data includes the steps of:
step S201: determining root nodes and leaf nodes of a tree forest based on the relation between the historical load data and preset cut point values; specifically, the tree forest construction process is as follows: taking sample data in the historical load data as a root node of each tree in the tree forest; placing any dimension data in a left leaf node or a right leaf node based on the size relation between any dimension data in the sample data and a preset cutting point value; and stopping generating new leaf nodes when each tree in the tree forest meets the preset condition.
When the obtained historical load data comprises load data of a plurality of days, load data of u days are randomly selected from the load data of a plurality of days, and u initial trees are constructed, wherein root nodes of the u initial trees are the load data of u days randomly selected from the historical load data respectively. Since load data of a plurality of times is included in load data of each day, the load data of each time can be regarded as one dimension. Then randomly choose one attribute dimension m, for example choose 9 per day: the load data acquired at the moment 15 is taken as the attribute dimension m, and a cut point value n (the requirement n is smaller than the maximum value and larger than the minimum value of the attribute dimension m in the current node) is randomly generated in all data in the current node, and the maximum value and the minimum value are specifically the maximum value and the minimum value in the acquired m-th load data of each day (such as the load data acquired at the moment 9:15 each day). And then dividing the data space of the current node into two subspaces by taking the cut point value n as a division plane. Placing data items with values smaller than n in a set formed by the mth load data of each day into a left leaf node of the current tree node; and otherwise, putting the node into the right leaf node. And repeatedly executing the process until only one load data (namely that the cutting cannot be continued) in the root node or the initial tree reaches the preset limit height, and obtaining the final tree forest.
Step S202: the height of each tree in the tree forest is calculated for each historical load data. After determining the tree forest, taking the obtained historical load data as test data x, bringing the test data into each tree of the tree forest, judging the position of the test data in the tree, determining the height of the test data falling on each tree, and marking the height as h (x), wherein the average value of all the heights of the trees corresponding to the test data is E (h (x)), and setting standard average search length as follows because the structure of an initial tree is similar to that of a binary search tree:
l(n)＝2H(n-1)-[2(n-1)/n] (1)
H(i)＝ln(i)+0.5772 (2)
where n is the cut point value, l (n) is the standard average search length, and H (i) is the harmonic number.
Step S203: calculating an anomaly score for each historical load data based on the average value of all heights and the preset cutting point data; wherein the anomaly score is calculated using the following formula:
u(x,n)＝2 E(h(x))l(n) (3)
wherein x is the data to be measured, n is the value of the cutting point, E (h (x)) is the average value of all depths h (x)
Step S204: determining whether the historical load data is abnormal data based on an abnormal score. Specifically, when the abnormality score is close to 1, the data is abnormal data. Thus, the anomaly score sum 1 can be made a difference, and when the difference is smaller than the threshold value, it is determined as anomaly data.
In one embodiment, the method for filling the missing data after deleting the abnormal data in the historical load data based on the KNN algorithm includes the following steps:
step S301: deleting the determined abnormal data to generate missing data; and after judging the abnormal data, deleting all the abnormal data, so as to convert the abnormal data into missing data.
Step S302: calculating the Euclidean distance of each missing data based on a data set matrix constructed by the historical load data; specifically, the dataset matrix is represented as: (x) 1 ,x 2 ,…,x n ) T Wherein [ X ]] y Is the y-th attribute of the data, y is less than or equal to w, n is the number of samples, and w is the data dimension. Wherein the data set matrix contains n samples (column vector, n=u if the data set matrix is constructed by load data of u days), and each sample contains w data (attributes); [ X ]] y Is the y-th data of a certain sample in the dataset matrix.
When the Euclidean distance is calculated, all the missing data are sequentially used as a missing example x iy Calculating the Euclidean distance:
wherein x is iy To miss an instance, x jy Is an undelayed example. Wherein x is i For missing data x iy Sample number, x where j For un-missing data x jy Sample number where; all the un-missing data and x are calculated jy The euclidean distance between the two is a euclidean distance corresponding to the missing data for each piece of data which is not missing.
Step S303: and calculating the substitution value of the missing data based on the k neighbor data selected by the Euclidean distance and the weight of the k neighbor data to fill. Specifically, after the euclidean distance is calculated, data records corresponding to K distances with the smallest distance are selected as K neighbor data of the target data. Then calculating weights of K nearest neighbor targets of the target data by adopting the following formula:
wherein d i For the ith adjacent point and the target point x iy Distance between them.
And (3) carrying out weighted average calculation through the calculated weight to serve as a substitution value of the missing data:
wherein x is ky Values representing nearest-neighbor corresponding properties, i.e. selected and x iy K data with minimum Euclidean distance between them, w ky Is a weighted average.
In one embodiment, determining a primary one of the weather factors associated with the baseline load prediction based on a principal component analysis includes the steps of:
step S401: carrying out standardization processing on meteorological factor data related to baseline load prediction; specifically, in order to eliminate the influence of dimension, the data is firstly standardized, and the processing formula is as follows:
y i,j ＝(x ij -x j )/S j (7)
wherein x is ij Partitioning the j index value for i in the original data; x is x j Is the sample mean value of the jth index in the original data, S j I=1, 2, …, n, j=1, 2, …, n, which is the sample standard deviation of the index. Wherein, the i partition refers to the i data of a certain index, and the index value refers to specific meteorological factor data of the index; in the present embodiment, j refers to any one of six indexes including barometric pressure, air temperature, humidity, wind speed, precipitation, and sunlight, and i refers to any one of 96 pieces of data (load data are acquired at intervals of 15 minutes) a day.
Step S402: calculating the characteristic value and the characteristic vector of the weather factor data after standardized processing; specifically, a normalized post-processing correlation matrix is constructed, and then a linear algebraic correlation knowledge is adopted to solve the eigenvalue and eigenvector of the correlation distance. Wherein KMO checksum butler verification is performed after determining the correlation matrix of the normalized data for verifying the correlation between the independent and dependent variables. Expressed as lambda for the calculated eigenvalues and eigenvectors, respectively 1 ≥λ 2 ≥…≥λ n And feature vector mu 1 ,μ 2 ,…,μ n These feature vectors are orthogonal in pairs.
Step S403: the primary weather factor is determined based on the principal component scores calculated using the eigenvectors and the cumulative contribution calculated using the eigenvalues. Wherein the principal component score is calculated using the following formula:
wherein k=1, 2, …, n, x j Is the sample mean value, mu of the j-th index in the original data j Is the corresponding feature vector.
And then calculating the accumulated contribution rate of each principal component to the baseline load prediction by adopting the calculated characteristic value, wherein the calculation formula is as follows:
where α is the number of principal components in analyzing the cumulative contribution rate, and λ is the eigenvalue. Wherein lambda is j Is one of all characteristic values lambda k Is selected from the first alpha eigenvalues of the accumulation.
As shown in fig. 2, when determining the main meteorological factors, the scores of the principal components and the principal components corresponding to the scores are arranged in a descending order, the first p (p < q) principal components with the accumulated contribution rate exceeding 80% are selected as reference factors for the subsequent baseline load prediction, q refers to the number of meteorological factor indexes, and in this embodiment, q=6.
In one embodiment, the baseline load prediction is performed based on a model constructed using the padded historical load data and the primary meteorological factor data, comprising the steps of:
step S501: decoupling the filled historical load data to obtain non-adjustable load data, and adjustable load data which does not contain air conditioner load; specifically, because the decoupling of the air conditioning load is inconvenient, the load data part structure is used for stripping the adjustable load (not including the air conditioning load) from the total load or all loads, so that the air conditioning load and the non-adjustable load are combined for load prediction. And when the electric meter is stripped, adding actual values of the electric meters at the distributed charging piles, the distributed energy storage and the like in commercial building users, wherein the sum of the actual values is the adjustable load actual value which does not contain the air conditioner load.
Step S502: and carrying out baseline load prediction by adopting a model constructed by non-adjustable load data, adjustable load data without air conditioner load, adjustable load data and main meteorological factor data. Specifically, the load prediction process is as follows: the method comprises the steps of constructing a proportion prediction model by taking total load data in historical loads, adjustable load data which does not contain air conditioning loads and main meteorological factor data as input and taking total load proportion occupied by the sum of the non-adjustable loads and the air conditioning loads as output; constructing a load prediction model by taking non-adjustable load, air conditioner load and main meteorological factor data as input and taking a baseline load value of the non-adjustable load and the air conditioner load as output; and determining the total baseline load based on the ratio of the baseline load predicted by the load prediction model and the proportion predicted by the proportion prediction model.
When the total load data, the adjustable load data without the air conditioner load and the main meteorological factor data in the historical load are used as input, the node number of the input layer of the neural network is 3, namely the total load data, the other adjustable load data except the air conditioner load and the main meteorological factor data are respectively used for predicting the total load proportion occupied by the sum of the non-adjustable load and the air conditioner load; therefore, the neural network is trained by input and output, and parameters of the neural network are adjusted, so that the prediction of the comparative example k can be realized.
Similarly, when the baseline load is predicted, the input is the non-adjustable load, the air conditioner load and the main meteorological factor data, the output is the baseline load of the non-adjustable load and the air conditioner load, the neural network is trained through the input and the output, the parameters of the neural network are adjusted, and the prediction of the baseline load Pb can be realized.
When the total baseline load of the predicted day is predicted, a typical day with the same attribute as the predicted day (the same working day or the same rest day) and no abnormal data, missing data or the repaired abnormal data and missing data is selected, the related data of the typical day are respectively input into two trained models, and the proportion of the sum of the non-adjustable load and the baseline load of the air conditioner load and the sum of the non-adjustable load and the air conditioner load of the predicted demand response day in the total load is respectively obtained, so that the calculation of the total baseline load is represented by adopting the following formula:
P＝P b /k (10)
wherein P is b For a baseline load value of the non-adjustable load and the air conditioning load for the predicted demand response day, k is the ratio of the non-adjustable load to the air conditioning load in the total load for the predicted demand response day.
In one embodiment, as shown in fig. 3, the weather factor-considered demand response baseline load prediction method is implemented by the following flow: acquiring load data and corresponding meteorological factor data; carrying out abnormal data identification and missing data filling on the load data; calculating meteorological factor data by adopting a principal component analysis method, and determining main meteorological factors; partially decoupling the load data and stripping the adjustable data; model training and prediction are carried out based on the decoupled load data and the determined main meteorological factors, wherein the trained model is mainly used for predicting the proportion k of the non-adjustable load and the air conditioner load in the total load and predicting the baseline load value of the non-adjustable load and the air conditioner load, the neural network parameters are adjusted through input and corresponding output during training, and the proportion of the demand response days and the baseline load value are input into the trained model through selected typical days with the same attribute as the predicted days during prediction; the baseline load on demand response days is then determined by the ratio of the two.
The embodiment of the invention also provides a device for predicting the demand response baseline load taking weather factors into consideration, as shown in fig. 4, which comprises:
the data acquisition module is used for acquiring historical load data; the specific content refers to the corresponding parts of the above method embodiments, and will not be described herein.
The abnormal data judging module is used for determining abnormal data in the historical load data based on the tree forest constructed by the historical load data; the specific content refers to the corresponding parts of the above method embodiments, and will not be described herein.
The filling module is used for filling the missing data after deleting the abnormal data in the historical load data based on a KNN algorithm; the specific content refers to the corresponding parts of the above method embodiments, and will not be described herein.
A principal component analysis module for determining a primary one of the weather factors associated with the baseline load prediction based on the principal component analysis; the specific content refers to the corresponding parts of the above method embodiments, and will not be described herein.
And the prediction module is used for predicting the baseline load based on a model constructed by adopting the filled historical load data and the main meteorological factor data. The specific content refers to the corresponding parts of the above method embodiments, and will not be described herein.
According to the device for predicting the demand response baseline load taking meteorological factors into consideration, historical load data are obtained; determining abnormal data in the historical load data based on the tree forest constructed by the historical load data; filling the missing data after deleting the abnormal data in the historical load data based on a KNN algorithm; determining a primary one of the weather factors associated with the baseline load prediction based on the principal component analysis; and carrying out baseline load prediction based on a model constructed by adopting filled historical load data and main meteorological factor data. Therefore, the recognition and filling of the abnormal data are realized through the constructed tree forest and KNN algorithm, so that the subsequent prediction result is more accurate; meanwhile, the influence of meteorological factors on load prediction is considered in the process of carrying out demand response baseline load prediction, so that the finally obtained baseline load prediction result is more accurate, and in addition, the factors with larger contribution rate in the meteorological factors affecting the load prediction are analyzed by using a principal component analysis method, so that the defect of slow algorithm calculation speed caused by considering all the meteorological factors is avoided.
The functional description of the weather factor-considered demand response baseline load prediction device provided by the embodiment of the invention is described in detail with reference to the weather factor-considered demand response baseline load prediction method in the above embodiment.
The embodiment of the present invention also provides a storage medium, as shown in fig. 5, on which is stored a computer program 601, which when executed by a processor, implements the steps of the demand response baseline load prediction method in the above embodiment, taking into account meteorological factors. The storage medium also stores audio and video stream data, characteristic frame data, interactive request signaling, encrypted data, preset data size and the like. The storage medium may be a magnetic Disk, an optical Disk, a Read-Only Memory (ROM), a Random access Memory (Random AccessMemory, RAM), a Flash Memory (Flash Memory), a Hard Disk (HDD), or a Solid State Drive (SSD); the storage medium may also comprise a combination of memories of the kind described above.
It will be appreciated by those skilled in the art that implementing all or part of the above-described embodiment method may be implemented by a computer program to instruct related hardware, where the program may be stored in a computer readable storage medium, and the program may include the above-described embodiment method when executed. Wherein the storage medium may be a magnetic Disk, an optical Disk, a Read-Only Memory (ROM), a random access Memory (RandomAccessMemory, RAM), a Flash Memory (Flash Memory), a Hard Disk (HDD), a Solid State Drive (SSD), or the like; the storage medium may also comprise a combination of memories of the kind described above.
The embodiment of the present invention further provides an electronic device, as shown in fig. 6, which may include a processor 51 and a memory 52, where the processor 51 and the memory 52 may be connected by a bus or other means, and in fig. 6, the connection is exemplified by a bus.
The processor 51 may be a central processing unit (Central Processing Unit, CPU). The processor 51 may also be other general purpose processors, digital signal processors (Digital Signal Processor, DSP), application specific integrated circuits (Application Specific Integrated Circuit, ASIC), field programmable gate arrays (Field-Programmable Gate Array, FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, or combinations thereof.
The memory 52 serves as a non-transitory computer readable storage medium that may be used to store non-transitory software programs, non-transitory computer-executable programs, and modules, such as corresponding program instructions/modules in embodiments of the present invention. The processor 51 executes various functional applications of the processor and data processing by running non-transitory software programs, instructions and modules stored in the memory 52 to implement the weather-factor-considered demand response baseline load prediction method in the above-described method embodiments.
The memory 52 may include a memory program area that may store an operating device, an application program required for at least one function, and a memory data area; the storage data area may store data created by the processor 51, etc. In addition, memory 52 may include high-speed random access memory, and may also include non-transitory memory, such as at least one magnetic disk storage device, flash memory device, or other non-transitory solid state storage device. In some embodiments, memory 52 may optionally include memory located remotely from processor 51, which may be connected to processor 51 via a network. Examples of such networks include, but are not limited to, the internet, intranets, local area networks, mobile communication networks, and combinations thereof.
The one or more modules are stored in the memory 52 and when executed by the processor 51 perform the weather factor-considered demand response baseline load prediction method of the embodiment shown in fig. 1-3.
The specific details of the electronic device may be understood in reference to the corresponding related descriptions and effects in the embodiments shown in fig. 1 to 3, which are not repeated herein.
Although embodiments of the present invention have been described in connection with the accompanying drawings, various modifications and variations may be made by those skilled in the art without departing from the spirit and scope of the invention, and such modifications and variations fall within the scope of the invention as defined by the appended claims.
Claims (10)
1. A method of demand response baseline load prediction taking into account meteorological factors, comprising:
acquiring historical load data;
determining abnormal data in the historical load data based on a tree forest constructed by the historical load data;
filling the missing data after deleting the abnormal data in the historical load data based on a KNN algorithm;
determining a primary one of the weather factors associated with the baseline load prediction based on the principal component analysis;
and carrying out baseline load prediction based on a model constructed by adopting filled historical load data and main meteorological factor data.
2. The method of claim 1, wherein determining anomaly data in the historical load data based on a tree forest constructed from the historical load data comprises:
determining root nodes and leaf nodes of a tree forest based on the relation between the historical load data and preset cut point values;
calculating the height of each tree in the tree forest of each historical load data;
calculating an anomaly score for each historical load data based on the average value of all heights and the preset cutting point data;
determining whether the historical load data is abnormal data based on an abnormal score.
3. The method of claim 2, wherein determining root and leaf nodes of a tree forest based on the historical load data and a relationship of the historical load data to a predetermined cut point value comprises:
taking sample data in the historical load data as a root node of each tree in the tree forest;
placing any dimension data in a left leaf node or a right leaf node based on the size relation between any dimension data in the sample data and a preset cutting point value;
and stopping generating new leaf nodes when each tree in the tree forest meets the preset condition.
4. The method for predicting a demand response baseline load taking meteorological factors into consideration according to claim 1, wherein filling missing data after deleting abnormal data in the historical load data based on a KNN algorithm comprises:
deleting the determined abnormal data to generate missing data;
calculating the Euclidean distance of each missing data based on a data set matrix constructed by the historical load data;
and calculating the substitution value of the missing data based on the k neighbor data selected by the Euclidean distance and the weight of the k neighbor data to fill.
5. The method of claim 1, wherein determining a primary one of the weather factors associated with the baseline load forecast based on a principal component analysis comprises:
carrying out standardization processing on meteorological factor data related to baseline load prediction;
calculating the characteristic value and the characteristic vector of the weather factor data after standardized processing;
the primary weather factor is determined based on the principal component scores calculated using the eigenvectors and the cumulative contribution calculated using the eigenvalues.
6. The method of claim 1, wherein the predicting the baseline load based on the model constructed using the filled historical load data and the primary weather factor data comprises:
decoupling the filled historical load data to obtain non-adjustable load data, and adjustable load data which does not contain air conditioner load;
and carrying out baseline load prediction by adopting a model constructed by non-adjustable load data, adjustable load data without air conditioner load, adjustable load data and main meteorological factor data.
7. The method of claim 6, wherein the baseline load prediction using a model constructed from non-adjustable load data, adjustable load data that does not include air conditioning load, adjustable load data, and primary weather factor data comprises:
the method comprises the steps of constructing a proportion prediction model by taking total load data in historical loads, adjustable load data which does not contain air conditioning loads and main meteorological factor data as input and taking total load proportion occupied by the sum of the non-adjustable loads and the air conditioning loads as output;
constructing a load prediction model by taking non-adjustable load, air conditioner load and main meteorological factor data as input and taking a baseline load value of the non-adjustable load and the air conditioner load as output;
and determining the total baseline load based on the ratio of the baseline load predicted by the load prediction model and the proportion predicted by the proportion prediction model.
8. A demand response baseline load prediction apparatus that accounts for meteorological factors, comprising:
the data acquisition module is used for acquiring historical load data;
the abnormal data judging module is used for determining abnormal data in the historical load data based on the tree forest constructed by the historical load data;
the filling module is used for filling the missing data after deleting the abnormal data in the historical load data based on a KNN algorithm;
a principal component analysis module for determining a primary one of the weather factors associated with the baseline load prediction based on the principal component analysis;
and the prediction module is used for predicting the baseline load based on a model constructed by adopting the filled historical load data and the main meteorological factor data.
9. A computer readable storage medium having stored thereon computer instructions for causing the computer to perform the weather factor considered demand response baseline load prediction method according to any one of claims 1-7.
10. An electronic device, comprising: a memory and a processor in communication with each other, the memory storing computer instructions, the processor executing the computer instructions to perform the weather-factor-considered demand response baseline load prediction method of any one of claims 1-7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211683644.6A CN116108974A (en) | 2022-12-27 | 2022-12-27 | Demand response baseline load prediction method and device considering meteorological factors |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211683644.6A CN116108974A (en) | 2022-12-27 | 2022-12-27 | Demand response baseline load prediction method and device considering meteorological factors |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN116108974A true CN116108974A (en) | 2023-05-12 |
Family
ID=86259021
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202211683644.6A Pending CN116108974A (en) | 2022-12-27 | 2022-12-27 | Demand response baseline load prediction method and device considering meteorological factors |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116108974A (en) |
-
2022
- 2022-12-27 CN CN202211683644.6A patent/CN116108974A/en active Pending
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109615129B (en) | Real estate customer transaction probability prediction method, server and computer storage medium | |
| CN107506868B (en) | Method and device for predicting short-time power load | |
| CN111625516A (en) | Method and device for detecting data state, computer equipment and storage medium | |
| CN111080117A (en) | Method and device for constructing equipment risk label, electronic equipment and storage medium | |
| CN109784528A (en) | Water quality prediction method and device based on time series and support vector regression | |
| CN113468034A (en) | Data quality evaluation method and device, storage medium and electronic equipment | |
| CN111598457B (en) | Method and device for determining quality of power wireless network | |
| CN113988441A (en) | Power wireless network link quality prediction and model training method and device | |
| CN111797320A (en) | Data processing method, device, equipment and storage medium | |
| CN114881343A (en) | Short-term load prediction method and device of power system based on feature selection | |
| CN110765351A (en) | Target user identification method and device, computer equipment and storage medium | |
| CN114595873A (en) | Grey-correlation-based DA-LSTM short-term power load prediction method | |
| CN113887866A (en) | Method and device for generating human living environment evaluation index | |
| CN111783883A (en) | Abnormal data detection method and device | |
| CN109615018B (en) | User personalized behavior evaluation method and device, computer equipment and storage medium | |
| CN111612648A (en) | Training method and device of photovoltaic power generation prediction model and computer equipment | |
| CN116680567A (en) | Missing state variable filling method and device for wind driven generator | |
| CN116431346A (en) | Compensation method for main memory capacity of electronic equipment | |
| CN116108974A (en) | Demand response baseline load prediction method and device considering meteorological factors | |
| CN113159537B (en) | Assessment method and device for new technical project of power grid and computer equipment | |
| CN111026661B (en) | Comprehensive testing method and system for software usability | |
| CN113469570A (en) | Information quality evaluation model construction method, device, equipment and storage medium | |
| CN112883267A (en) | Data heat degree statistical method and device based on deep learning | |
| CN114092216A (en) | Enterprise credit rating method, apparatus, computer device and storage medium | |
| CN113537693A (en) | Personnel risk level obtaining method, terminal and storage device |
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 |