Photovoltaic user baseline load estimation method, device and terminal equipment

Abstract

The embodiments of the present invention relate to the technical field of power system demand response, and disclose a photovoltaic user baseline load estimation method, device and terminal equipment. The above-mentioned photovoltaic user baseline load estimation method includes: obtaining the net load power and temperature data of all user clusters, and the observable photovoltaic output data of some photovoltaic users throughout the year; clustering the observable photovoltaic output data to obtain the generalized weather type label of each day; dividing all user clusters into photovoltaic user clusters and non-photovoltaic user clusters according to the weather type label; determining the first baseline load estimation value of the photovoltaic user cluster according to multiple photovoltaic user net load power sets; establishing a piecewise linear regression model according to the net load power and temperature data of the non-photovoltaic user cluster, and determining the second baseline load estimation value of the non-photovoltaic user cluster according to the piecewise linear regression model; determining the baseline load estimation value of all user clusters according to the first baseline load estimation value and the second baseline load estimation value.

Description

Technical Field

[0001] The present invention relates to the technical field of power system demand response, and in particular to a photovoltaic user baseline load estimation method, device and terminal equipment. Background Art

[0002] Demand response uses price and incentive signals to achieve massive aggregation and rapid response of demand-side resources, playing an important role in improving the system's supply and demand balance capabilities, shaving peak loads and filling valleys, and promoting the consumption of new energy. The user cluster baseline load is the basis for the recognition of the demand response effect between the system operator and the load aggregator. An overestimation or underestimation of the user cluster baseline load will affect the recognition of the demand response effect. Therefore, accurate user cluster baseline load estimation is very important for the implementation of demand response.

[0003] In recent years, the distributed photovoltaic industry has developed rapidly, and more and more residential users have begun to install distributed photovoltaic power generation systems. Most small distributed photovoltaic systems (<10kWp) are installed behind the meter. The user's meter measures the net load, that is, the user's load minus the photovoltaic output, while the system operator and load aggregator cannot observe the photovoltaic output. This unobservable photovoltaic output greatly increases the difficulty of estimating the baseline load. There are obvious differences in photovoltaic output under different weather types, which will lead to large differences in the net load curve of photovoltaic users with distributed photovoltaics installed in different weather types, while the net load curve of non-photovoltaic users without distributed photovoltaics is relatively similar, that is, the net load of photovoltaic users is more easily affected by weather factors than non-photovoltaic users. How to distinguish these two types of users and accurately estimate the baseline load has become a major problem in the era of distributed power grid connection. With the development of theoretical technologies such as artificial intelligence and machine learning, intelligent algorithms have been applied in the field of baseline load estimation, providing a guarantee for the implementation of distributed photovoltaic power grid demand response. Summary of the invention

[0004] In view of this, an embodiment of the present invention provides a method for estimating a baseline load of a photovoltaic user to solve the problem of low accuracy of baseline load estimation in the prior art.

[0005] In order to achieve the above object, the present invention adopts the following technical scheme:

[0006] In a first aspect, an embodiment of the present invention provides a method for estimating a baseline load of a photovoltaic user, comprising: obtaining net load power and temperature data of all user clusters, and observable photovoltaic output data of some photovoltaic users throughout the year; clustering the observable photovoltaic output data throughout the year to obtain multiple photovoltaic output power sets, setting generalized weather type labels for the multiple photovoltaic output power sets, and obtaining generalized weather type labels for each day of the year; dividing the entire user cluster into a photovoltaic user cluster and a non-photovoltaic user cluster according to the generalized weather type labels; dividing the net load power of the photovoltaic user cluster into multiple net load power sets according to the generalized weather type labels, establishing support vector machine regression models respectively according to the multiple net load power sets, and determining a first baseline load estimate; establishing a piecewise linear regression model according to the net load power and temperature data of the non-photovoltaic user cluster, and determining a second baseline load estimate according to the piecewise linear regression model; determining a baseline load estimate for all user clusters according to the first baseline load estimate and the second baseline load estimate.

[0007] In an embodiment of the present invention, a weather type classification method that relies only on photovoltaic output and not on external weather conditions is proposed. Through this method, the weather throughout the year can be classified according to generalized weather types, and then the generalized weather type labels for each day of the year are obtained. Through the obtained generalized weather type labels, when estimating the baseline load of the photovoltaic user cluster, different generalized weather types can be modeled separately. Compared with directly estimating the baseline load, the present invention can reasonably consider the impact of weather factors on photovoltaic output and net load, and the estimation results of classification modeling are more accurate.

[0008] Based on the first aspect, in some embodiments, the observable photovoltaic output data throughout the year is clustered to obtain multiple photovoltaic output power sets, and generalized weather type labels for the multiple photovoltaic output power sets are set to obtain generalized weather type labels for each day of the year, including: extracting the observable photovoltaic output data of the part of photovoltaic users throughout the year to obtain a set of photovoltaic output curves throughout the year; clustering the set of photovoltaic output curves throughout the year to obtain multiple photovoltaic output power sets, and calculating the average photovoltaic output of the multiple photovoltaic output power sets; each photovoltaic output power set contains multiple photovoltaic output curves; sorting the multiple photovoltaic output power sets according to the average photovoltaic output at typical photovoltaic power generation moments, and setting weather labels for the multiple photovoltaic output power sets according to the sorting results to obtain generalized weather type labels for each day of the year.

[0009] Based on the first aspect, in some embodiments, dividing the entire user cluster into a photovoltaic user cluster and a non-photovoltaic user cluster according to the generalized weather type label includes: extracting the average net load power of each user in the entire user cluster under different generalized weather type labels; for the entire user cluster, extracting the first feature F1 and the second feature F2, and combining the first feature F1 and the second feature F2 into a feature vector F n =[F1,F2] T ; According to the feature vector F n , dividing the entire user cluster into the photovoltaic user cluster and the non-photovoltaic user cluster.

[0010] Based on the first aspect, in some embodiments, the first feature F1 is expressed as follows: in, represents the average net load power of user n at sampling point t under the D-th weather condition, It represents the average net load power of user n at sampling point t under Class A weather conditions.

[0011] Based on the first aspect, in some embodiments, the process of determining the second feature F2 is: according to the preset time window α=[t s ,t e ], there is a connection starting point t s and end point t e The sampling points y(t) of each time period on the straight line are: Where, t s and t e is the sampling point in the determined time window α, represents the average net load power of user n at sampling point t under the A-th weather condition. The sampling points y(t) in each time period on the straight line constitute a set S, S = {y(t)|t∈[t s ,t e ]}. The set of sampling points located below the straight line S It is expressed as: The calculation formula of the second feature F2 is: F2 = card ( S ) / card(S), where card(·) represents the number of elements in the set.

[0012] Based on the first aspect, in some embodiments, a support vector machine regression model is established according to the multiple net load power sets, and a first baseline load estimate value of a photovoltaic user cluster is determined according to the support vector machine regression model, including: accumulating the net load power of each photovoltaic user in the photovoltaic user cluster to obtain the net load power of the photovoltaic user cluster; according to the generalized weather type label, classifying the net load power of the photovoltaic user cluster by label to obtain multiple net load power sets corresponding to each of the generalized weather type labels; establishing support vector machine regression models for the multiple net load powers respectively, and determining the first load estimate value of the photovoltaic user cluster according to the support vector machine regression model.

[0013] Based on the first aspect, in some embodiments, the step of establishing a piecewise linear regression model based on the net load power and temperature data of the non-photovoltaic user cluster and determining the second baseline load estimate value based on the piecewise linear regression model includes: accumulating the first net load power of each non-photovoltaic user in the non-photovoltaic user cluster to obtain the second net load power of the non-photovoltaic user cluster; the first net load power is the net load power of the non-photovoltaic user at the sampling point, and the second net load power is the net load power of the non-photovoltaic user cluster at the sampling point; the sampling point is a pre-divided time node; according to the second net load power, determining the net load power of the non-photovoltaic user cluster within a preset time period; segmenting the outdoor temperature data samples of multiple sampling points within the preset time period to obtain segmented outdoor temperature sample data; establishing a piecewise linear regression model corresponding to the preset time period according to the preset time period and the segmented outdoor temperature sample data; obtaining the first outdoor temperature collected at each sampling point, the first outdoor temperature being the outdoor temperature collected at each time sampling point in the target time period; segmenting the first outdoor temperature to obtain the second outdoor temperature, and inputting the second outdoor temperature into the piecewise linear regression model corresponding to the target time period to obtain the second baseline load estimate value.

[0014] In a second aspect, an embodiment of the present invention provides a photovoltaic user baseline load estimation device, comprising: an information acquisition module, for acquiring the net load power and temperature data of the entire user cluster, and the observable photovoltaic output data of some photovoltaic users throughout the year;

[0015] A classification module is used to cluster the photovoltaic output data throughout the year to obtain multiple photovoltaic output power sets, set generalized weather type labels for the multiple photovoltaic output power sets, and obtain generalized weather type labels for each day of the year; the entire user cluster is divided into a photovoltaic user cluster and a non-photovoltaic user cluster according to the generalized weather type labels; a first load estimation module is used to divide the net load power of the photovoltaic user cluster into multiple net load power sets according to the generalized weather type labels, and establish support vector machine regression models according to the multiple net load power sets to determine a first baseline load estimate; a second load estimation module is used to establish a piecewise linear regression model according to the net load power and temperature data of the non-photovoltaic user cluster, and determine a second baseline load estimate according to the piecewise linear regression model; a baseline load determination module is used to determine the baseline load estimate of the entire user cluster according to the first baseline load estimate and the second baseline load estimate.

[0016] In a third aspect, an embodiment of the present invention provides a terminal device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the computer program, the steps of the photovoltaic user baseline load estimation method as described in any one of the first aspects above are implemented.

[0017] In a fourth aspect, an embodiment of the present invention provides a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps of the photovoltaic user baseline load estimation method as described in any one of the above-mentioned first aspects are implemented.

[0018] In the embodiment of the present invention, the net power load of all users is directly used for classification and estimation, which solves the problem of low accuracy of baseline load estimation of user clusters under unobservable photovoltaic output, and helps to ensure the smooth implementation of power system demand response. Since photovoltaic users and non-photovoltaic users are affected by weather factors to different degrees, classification considerations can be modeled and analyzed according to their respective characteristics, which helps to improve the accuracy of baseline load estimation of user clusters, thereby accurately calculating the load reduction of user clusters during the execution of demand response, which can effectively improve the fairness of both parties involved in incentive-based demand response and help the implementation and promotion of demand response. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative labor.

[0020] Figure 1It is a flow chart of a photovoltaic user baseline load estimation method provided by an embodiment of the present invention;

[0021] Figure 2 is a flowchart of step 102 provided in an embodiment of the present invention;

[0022] Figure 3 is a flowchart of step 103 provided in an embodiment of the present invention;

[0023] Figure 4 is a flowchart of step 104 provided in an embodiment of the present invention;

[0024] Figure 5 It is a schematic diagram of the structure of a photovoltaic user baseline load estimation device provided by an embodiment of the present invention;

[0025] Figure 6 It is a schematic diagram of a terminal device provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0026] The present invention will be described more clearly below in conjunction with specific embodiments. The following embodiments will help those skilled in the art to further understand the effects of the present invention, but are not intended to limit the present invention in any form. It should be noted that, for those of ordinary skill in the art, several variations and improvements may be made without departing from the concept of the present invention. These all belong to the protection scope of the present invention.

[0027] In order to make the purpose, technical solutions and advantages of the present invention more clear, specific embodiments will be described below in conjunction with the accompanying drawings.

[0028] The embodiment of the present invention provides a method for estimating a baseline load of a photovoltaic user, such as Figure 1 As shown, the photovoltaic user baseline load estimation method may include steps 101 to 105.

[0029] Step 101: Obtain the net load power and temperature data of all users, and the observable photovoltaic output data of some photovoltaic users throughout the year.

[0030] Photovoltaic output refers to the active output power of photovoltaic power generation equipment, and the annual photovoltaic output is the photovoltaic output power of the user throughout the year. System operators can obtain observable photovoltaic output data of some users, and analyze the weather type throughout the year based on the photovoltaic output power of some users. However, the photovoltaic output data of most users are unknown to system operators. This is mainly due to the mechanism of self-generation and self-use by users in the distributed photovoltaic power generation area. Therefore, the present invention chooses to estimate the baseline load of the user cluster through the net load power of the user. The net load power is the data measured by the user's meter. Most photovoltaic systems are installed behind the meter, and the data measured by the user's meter is the user's load minus the photovoltaic output. At the same time, due to the strong correlation between distributed photovoltaic power generation output and weather, it is also necessary to obtain the annual temperature data of the area.

[0031] Step 102: Cluster the observable photovoltaic output data of some photovoltaic users throughout the year to obtain multiple photovoltaic output power sets, set generalized weather type labels for the multiple photovoltaic output power sets, obtain generalized weather type labels for each day of the year, and divide all user clusters into photovoltaic user clusters and non-photovoltaic user clusters according to the generalized weather type labels.

[0032] In a specific implementation, step 102 may include step 1021 to step 1022 .

[0033] Photovoltaic output is closely related to weather types. Photovoltaic output varies significantly under different weather types. The present invention does not rely on external weather conditions, but only relies on analyzing the photovoltaic output curve. First, the generalized weather type labels for each day of the year are determined, and then user clusters are classified according to the weather type labels to identify photovoltaic users and non-photovoltaic users in the user clusters.

[0034] Step 1021: clustering the photovoltaic outputs of all photovoltaic user clusters to obtain multiple photovoltaic output power sets, setting generalized weather type labels for the multiple photovoltaic output power sets, and obtaining generalized weather type labels for each day of the year.

[0035] In a specific implementation, step 1021 may include steps 10211 to 10213 .

[0036] Step 10211: extract the observable photovoltaic output data of some photovoltaic users throughout the year to obtain a set of photovoltaic output curves throughout the year.

[0037] Extract the photovoltaic output data of M photovoltaic users for X days throughout the year, calculate the average photovoltaic output value of each sampling point in a day, and the expression of the average photovoltaic output of M photovoltaic users at sampling point t on the xth day is:

[0038]

[0039] where px,m (t) represents the PV output of user m (m=1,2,…,M) at sampling point t on the xth (x=1,2,…,X) day, It represents the average photovoltaic output of sampling point t on the xth day. In the present invention, all sampling points t with the same value correspond to the same time, t=1,...,δ, δ is the number of load sampling points in a day, for example, if the load data is sampled once every hour, then δ=24.

[0040] According to the average photovoltaic output value at each sampling point t on day x The average photovoltaic output values ​​of each sampling point in a day are combined to obtain the average photovoltaic output curve of the day Then we can get the average photovoltaic output curve set for the whole year.

[0041] Step 10212: clustering the photovoltaic output curve set for the whole year to obtain multiple photovoltaic output power sets, and calculating the average photovoltaic output of the multiple photovoltaic output power sets, each photovoltaic output power set containing multiple photovoltaic output curves.

[0042] The K-means algorithm is used to cluster the annual average photovoltaic output curve set P to obtain multiple subsets, and the average photovoltaic output of the subsets at each sampling point t is calculated.

[0043] For example, the number of clusters is set to 4, and the subset {C1,...,C4} is obtained. Taking the subset C1 as an example, assuming that after clustering, C1 contains the average photovoltaic output curves of L days, the average photovoltaic output of all curves in the subset C1 at sampling point t can be obtained by the following expression:

[0044]

[0045] In the formula, represents the average photovoltaic output value of the lth (l=1,2,…,L) curve in the subset C1 at the sampling point t, It represents the average photovoltaic output of all curves in subset C1 at sampling point t. The average photovoltaic output of all sampling points constitutes the average photovoltaic output curve of subset C1. Similarly, the average photovoltaic output of all curves in subsets C2, C3, and C4 can also be obtained.

[0046] Step 10213: sorting the photovoltaic output average values ​​of the multiple photovoltaic output power sets at the typical photovoltaic power generation time, and setting the weather labels of the multiple photovoltaic output power sets according to the sorting results to obtain the generalized weather type label for each day of the year.

[0047] The average photovoltaic output at the time of photovoltaic power generation in the subsets is sorted, and the weather labels of each subset are set according to the sorting results. For example, at 12 o'clock, the sunlight is strong and the photovoltaic output is high, which can be used as a typical photovoltaic power generation time. The average photovoltaic output of the four subsets when the sampling point t is 12 o'clock is taken, sorted from high to low, and the weather type labels are set in the order of ABCD. The weather type in the subset with the highest average value is set to label A, and the weather type in the subset with the lowest average value is set to label D, and the generalized weather type labels for each day of the year are obtained.

[0048] Step 1022: Divide all user clusters into photovoltaic user clusters and non-photovoltaic user clusters according to the generalized weather type labels.

[0049] In step 1021, clustering is performed to obtain generalized weather type labels for each day of the year. Two features are defined using the weather type labels to analyze the net load power characteristics of all users in the user cluster, and feature vectors are extracted for classification, so that the user cluster is divided into two sub-clusters: a photovoltaic user cluster and a non-photovoltaic user cluster.

[0050] In a specific implementation, step 1022 may include steps 10221 to 10223 .

[0051] Step 10221: extract the average net load power of each user in the entire user cluster under different generalized weather types.

[0052] For a user cluster containing N users, the average net load power of user n (n=1,…,N) at sampling point t under different generalized weather types is extracted. For example, the average net load power expression of user n at sampling point t under the jth, j∈{A,B,C,D} type of weather can be:

[0053]

[0054] In the formula, Indicates that user n is in the lth (l=1,2,…,L j ) is the net load power at sampling point t on day 1, where the superscript net represents the net load power; represents the average net load power of user n at sampling point t under the jth, j∈{A,B,C,D} weather type; L j Indicates the number of days included in the jth weather category.

[0055] Step 10222: Extract the first feature F1 and the second feature F2 from all user clusters, and combine the first feature F1 and the second feature F2 into a feature vector F n =[F1,F2] T .

[0056] The first feature F1 and the second feature F2 are used to characterize the net load power difference between the photovoltaic user cluster and the non-photovoltaic user cluster. First, a time window α = [t s ,t e ], t s represents the starting time, t e Indicates the end time, based on the user's net load power sampled at different time points within the time window.

[0057] (1) The ratio of the absolute values ​​of the net load of user n under Class D and Class A weather conditions is taken as the first feature, denoted as F1.

[0058] The expression of F1 is as follows:

[0059]

[0060] in, represents the average net load power of user n at sampling point t under the D-th weather condition, It represents the average net load power of user n at sampling point t under Class A weather. The ratio F1 between the two can reflect the correlation between weather and net load power of different users. For photovoltaic users, the value of this feature should be greater than 1; for non-photovoltaic users, the value of this feature should be close to 1. If the value of F1 defined in the present invention is greater than 1.2, the user is a photovoltaic user; if the value of F1 is in the interval [0.95,1.05], it is a non-photovoltaic user.

[0061] (2) Based on the average net load curve under the generalized weather type A, the second feature is extracted, called the “concave-convex characteristic”, denoted as F2.

[0062] According to the determined time window α = [t s ,t e ], there is a line connecting the starting point t s and end point t e The expression of the straight line is as follows:

[0063]

[0064] In the formula, y(t) represents the sampling points at each time period on the straight line, t s and t e is the sampling point in the determined time window α, represents the average net load power of user n at sampling point t under Class A weather conditions; all sampling points constitute a set, denoted as S, S = {y(t)|t∈[t s ,t e ]}.

[0065] The set of sampling points below the straight line isS It can be expressed as:

[0066]

[0067] in,

[0068] The calculation formula for the second feature F2 is obtained as follows:

[0069] F2=card( S ) / card(S) (7)

[0070] In the formula, card(·) represents the number of elements in the set. Feature F2 can be used to reflect the concave and convex characteristics of the average net load curve within the time window δ. In Class A weather, the net load curve of photovoltaic users will have a significant concave. Taking 11:00 to 14:00 in a day as an example, the values ​​corresponding to these two points on the average net load curve are connected into a straight line, and the curve will appear below the straight line; while for non-photovoltaic users, the net load curve and the connected straight line will alternate up and down, and even the net load curve will be completely above the connected straight line. The F2 feature describes the proportion of sampling points below this line. If the value of F2 is between the interval [0.5,1], it means that the user is a photovoltaic user; if the value of F2 is between the interval [0,0.5), it means that the user is a non-photovoltaic user.

[0071] For all user clusters, extract features F1 and F2, and the features of user n can form a feature vector F n =[F1,F2] T .

[0072] Step 10223: According to the feature vector F n Classify all user clusters into photovoltaic user clusters and non-photovoltaic user clusters.

[0073] The normalized N user feature vector set F = {F1,...,F N} as input, the K-means algorithm is used to cluster all users into two categories. The category with the larger eigenvalue of the cluster center corresponds to photovoltaic users, and the rest are non-photovoltaic users. Thus, a photovoltaic user cluster containing N1 users and a non-photovoltaic user cluster containing N2 users are obtained, and N1+N2=N.

[0074] Step 103: Divide the net load power of the photovoltaic user cluster into multiple net load power sets according to the generalized weather type label, establish support vector machine regression models according to the multiple net load power sets, and determine a first baseline load estimation value.

[0075] In a specific implementation, step 103 may include steps 1031 to 1034 .

[0076] Step 1031: The net load power of each photovoltaic user in the photovoltaic user cluster is accumulated to obtain the net load power of the photovoltaic user cluster.

[0077] The net load power expression is as follows:

[0078]

[0079] In the formula, represents the net load power of the nth (n=1,...,N1, N1 is the total number of users in the cluster) photovoltaic user at sampling point t, It represents the net load power of the PV user cluster at sampling point t on day d. The net load power of each sampling point is connected to form the daily net load power.

[0080] Step 1032: According to the four generalized weather type labels {A, B, C, D}, the annual daily net load power of the photovoltaic user cluster is divided into four categories according to the labels.

[0081] Step 1033: Establish support vector machine regression models for the net load powers of user clusters with different weather type labels.

[0082] For example, a set of net load power data for R days is included. The average net load power data for the DR days is removed, and the model is trained using the data of the remaining non-DR days. DR represents the implementation of demand response, where the DR day represents a day on which demand response is implemented, and the DR period represents a period of time during which demand response is implemented.

[0083] When training the support vector machine regression model, the input and output of the model must be determined. In this step, the input of the model is set to the load data of the 11 sampling points before sampling point t on day d. Temperature data of the first five sampling points The output is the load at sampling point t on day d The trained model is recorded as model.

[0084] Step 1034: Estimate the baseline load of the photovoltaic user cluster according to the support vector machine regression model, and determine a first load estimation value of the photovoltaic user cluster.

[0085] First, determine the weather type of the DR day, and then call the model trained for this type of weather type set, input the load and temperature data of each sampling point before the DR period of the day, and you can get the baseline load of the PV user cluster during the DR period on the DR day. The baseline load value of the PV user cluster is the first load estimate.

[0086] Step 104: Establish a piecewise linear regression model based on the net load power and temperature data of the non-photovoltaic user cluster, and determine a second baseline load estimation value based on the piecewise linear regression model.

[0087] In a specific implementation, the process of step 104 may include steps 1041 to 1045 .

[0088] Step 1041: Accumulate the first net load power of each non-photovoltaic user in the non-photovoltaic user cluster to obtain the second net load power of the non-photovoltaic user cluster.

[0089] The net load power expression is as follows:

[0090]

[0091] In the formula, represents the first net load power of the nth non-PV user at sampling point t on the dth day, represents the second net load power of the non-PV user cluster at sampling point t on day d.

[0092] Step 1042: Determine the net load power of the non-photovoltaic user cluster within a preset time period according to the second net load power.

[0093] For example, the modeling analysis is performed on a weekly basis, and the net load power from Monday to Sunday is determined based on the net load power value of each sampling point. is the net load power of the i-th sampling point of the non-PV user cluster in a week, i=1,2,...β, where β=7*δ, δ is the number of load sampling points in a day. For example, if the load data is sampled once every hour, there are 7*24=168 sampling points in a week. Using the 24-hour system, if the first sampling point is 0:00 on Monday, the 25th sampling point is 0:00 on Tuesday, and the 168th sampling point is 11:00 pm on Sunday (23:00).

[0094] Step 1043: segment the outdoor temperature data samples of multiple sampling points within a preset time period to obtain segmented outdoor temperature sample data.

[0095] For example, the outdoor temperature data of each sampling point i can be divided into 6 segments. Before segmenting, several expressions are defined. T(i) represents the temperature of sampling point i (i = 1, 2, ... β) within a week; T j (i) represents the temperature of the jth segment after the temperature of sampling point i is segmented, j = 1, 2, ... 6. k (k=1,2,...5) represents the boundary of the temperature interval. The segment temperature calculation method is as follows:

[0096] If T(i)≤B1, then T1(i)=T(i), Ta (i) = 0 (a = 2, 3, ... 6);

[0097] If T(i)>B1, then T1(i)=B1;

[0098] If T(i)>B b (b=1,2,3,4), then T b (i) = B b+1 -B b , otherwise T b (i) = T(i) - B b-1 And T p (i) = 0 (p = b + 1, ... 6);

[0099] If T(i)>B5, then T5(i)=B5-B4 and T6(i)=T(i)-B5.

[0100] An example of the temperature segmentation method (assuming that the boundaries of the temperature intervals from B1 to B5 are 10, 20, ..., 50 respectively) is shown in Table 1.

[0101] Table 1 Example of temperature segment calculation

[0102]

[0103] Step 1044: According to the preset time period and the segmented outdoor temperature sample data, a segmented linear regression model corresponding to the preset time period is established.

[0104] For example, a piecewise linear regression model containing a week's time variable and piecewise temperature data can be established. The piecewise temperature data of each sampling point i in a week is obtained. This data is used as the input of the model, and the net load of each sampling point i in a week is used as the output of the model. The model is established as follows:

[0105]

[0106] In the formula, α i is the regression parameter corresponding to sampling point i, β j is the regression parameter corresponding to the temperature of the jth segment at sampling point i, is the net load power of the non-PV user cluster at sampling point i, α i and β j It can be obtained by the least square method. Thus, the piecewise linear regression model of each sampling point in a week is obtained.

[0107] Step 1045: Obtain the first outdoor temperature collected at each sampling point, where the first outdoor temperature is the outdoor temperature collected at each time sampling point in the target time period, segment the first outdoor temperature, obtain the second outdoor temperature, input the second outdoor temperature into the segmented linear regression model corresponding to the target time period, and obtain the second baseline load estimate.

[0108] For example, the first outdoor temperature collected at each sampling point in the time period to be calculated can be obtained first, and then the time position of the time period in a week can be determined. The piecewise linear regression model of the corresponding time period can be called, and the first outdoor temperature data of the sampling points in the time period can be segmented and then input into the piecewise linear regression model to estimate the baseline load of the time period to be calculated, and obtain the second baseline load value of non-photovoltaic users.

[0109] Step 105: Add the baseline load estimation values ​​of the photovoltaic users and the baseline load estimation values ​​of the non-photovoltaic users to obtain the baseline load estimation values ​​of the entire user cluster.

[0110] The estimated value of the user cluster baseline load is expressed as follows:

[0111]

[0112] In the formula, represents the estimated net load power of the PV user cluster at sampling point t on day d, represents the estimated net load power of the non-PV user cluster at sampling point t on day d, It represents the estimated net load power of the user cluster at sampling point t on the dth day, d represents a DR day, and t represents a moment in the DR period of the DR day.

[0113] See also Figure 5 The embodiment of the present invention provides a photovoltaic user baseline load estimation device 30, including: an information acquisition module 310, a classification module 320, a first load estimation module 330, a second load estimation module 340 and a baseline load estimation module 350.

[0114] The information acquisition module 310 is used to obtain the net load power and temperature data of all user clusters, as well as the observable photovoltaic output data of some photovoltaic users throughout the year.

[0115] The classification module 320 clusters the observable photovoltaic output data throughout the year to obtain multiple photovoltaic output power sets, sets generalized weather type labels for the multiple photovoltaic output power sets, and obtains generalized weather type labels for each day of the year; and divides all user clusters into photovoltaic user clusters and non-photovoltaic user clusters according to the generalized weather type labels.

[0116] The first load estimation module 330 is used to divide the net load power of the photovoltaic user cluster into multiple net load power sets according to the generalized weather type label, establish support vector machine regression models according to the multiple net load power sets, and determine a first baseline load estimation value.

[0117] The second load estimation module 340 is used to establish a piecewise linear regression model according to the net load power and temperature data of the non-photovoltaic user cluster, and determine a second baseline load estimation value according to the piecewise linear regression model.

[0118] The baseline load determination module 350 is used to determine the baseline load estimation value of the entire user cluster according to the first baseline load estimation value and the second baseline load estimation value.

[0119] Figure 6 Schematic diagram of a terminal device provided by an embodiment of the present invention. Figure 6 As shown, the terminal device 4 of this embodiment includes: a processor 40, a memory 41, and a computer program 42 stored in the memory 41 and executable on the processor 40, such as a baseline load estimation program. When the processor 40 executes the computer program 42, the steps in the above-mentioned photovoltaic user baseline load estimation method embodiment are implemented, such as Figure 1 Alternatively, when the processor 40 executes the computer program 42, the functions of each module / unit in the above-mentioned device embodiments are realized, for example, Figure 5 The functions of modules 310 to 350 are shown.

[0120] Exemplarily, the computer program 42 may be divided into one or more modules / units, which are stored in the memory 41 and executed by the processor 40 to implement the present invention. The one or more modules / units may be a series of computer program instruction segments capable of implementing specific functions, which are used to describe the execution process of the computer program 42 in the terminal device 4. For example, the computer program 42 may be divided into an information acquisition module, a classification module, a first load estimation module, a first load estimation module, and a baseline load estimation module.

[0121] The terminal device 4 may be a computing device such as a desktop computer, a notebook, a PDA, or a cloud server. The terminal device may include, but is not limited to, a processor 40 and a memory 41. Those skilled in the art will appreciate that Figure 6 It is only an example of the terminal device 4 and does not constitute a limitation on the terminal device 4. It may include more or fewer components than shown in the figure, or a combination of certain components, or different components. For example, the terminal device may also include input and output devices, network access devices, buses, etc.

[0122] The processor 40 may be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor may be a microprocessor or any conventional processor, etc.

[0123] The memory 41 may be an internal storage unit of the terminal device 4, such as a hard disk or memory of the terminal device 4. The memory 41 may also be an external storage device of the terminal device 4, such as a plug-in hard disk, a smart media card (SMC), a secure digital (SD) card, a flash card, etc. equipped on the terminal device 4. Further, the memory 41 may also include both an internal storage unit and an external storage device of the terminal device 4. The memory 41 is used to store the computer program and other programs and data required by the terminal device. The memory 41 may also be used to temporarily store data that has been output or is to be output.

[0124] Those skilled in the art can clearly understand that for the convenience and simplicity of description, only the division of the above-mentioned functional units and modules is used as an example for illustration. In practical applications, the above-mentioned function allocation can be completed by different functional units and modules as needed, that is, the internal structure of the device can be divided into different functional units or modules to complete all or part of the functions described above. The functional units and modules in the embodiment can be integrated into a processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The above-mentioned integrated unit can be implemented in the form of hardware or in the form of software functional units. In addition, the specific names of the functional units and modules are only for the convenience of distinguishing each other, and are not used to limit the scope of protection of the present invention. The specific working process of the units and modules in the above-mentioned system can refer to the corresponding process in the aforementioned method embodiment, which will not be repeated here.

[0125] In the above embodiments, the description of each embodiment has its own emphasis. For parts that are not described or recorded in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0126] Those skilled in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of the present invention.

[0127] In the embodiments provided by the present invention, it should be understood that the disclosed devices / terminal equipment and methods can be implemented in other ways. For example, the device / terminal equipment embodiments described above are only schematic. For example, the division of the modules or units is only a logical function division. There may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms.

[0128] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0129] In addition, each functional unit in each embodiment of the present invention may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of software functional units.

[0130] If the integrated module / unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the present invention implements all or part of the processes in the above-mentioned embodiment method, and can also be completed by instructing the relevant hardware through a computer program. The computer program can be stored in a computer-readable storage medium, and the computer program can implement the steps of the above-mentioned various method embodiments when executed by the processor. Among them, the computer program includes computer program code, and the computer program code can be in source code form, object code form, executable file or some intermediate form. The computer-readable medium may include: any entity or device capable of carrying the computer program code, recording medium, U disk, mobile hard disk, disk, optical disk, computer memory, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), electric carrier signal, telecommunication signal and software distribution medium. It should be noted that the content contained in the computer-readable medium can be appropriately increased or decreased according to the requirements of legislation and patent practice in the jurisdiction. For example, in some jurisdictions, according to legislation and patent practice, computer-readable media do not include electric carrier signals and telecommunication signals.

[0131] The embodiments described above are only used to illustrate the technical solutions of the present invention, rather than to limit the same. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that the technical solutions described in the aforementioned embodiments may still be modified, or some of the technical features may be replaced by equivalents. Such modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention, and should all be included in the protection scope of the present invention.

Claims

1. A photovoltaic user baseline load estimation method, characterized in that: include: Obtain the net load power and temperature data of all users, as well as the observable photovoltaic output data of some photovoltaic users throughout the year; Clustering the observable photovoltaic output data to obtain multiple photovoltaic output power sets, setting generalized weather type labels for the multiple photovoltaic output power sets, and obtaining generalized weather type labels for each day of the year; dividing the entire user cluster into a photovoltaic user cluster and a non-photovoltaic user cluster according to the generalized weather type labels; dividing the net load power of the photovoltaic user cluster into multiple net load power sets according to the generalized weather type labels, establishing support vector machine regression models based on the multiple net load power sets, and determining a first baseline load estimate; establishing a piecewise linear regression model based on the net load power and temperature data of the non-photovoltaic user cluster, and determining a second baseline load estimate based on the piecewise linear regression model; Determining a baseline load estimate for the entire user cluster based on the first baseline load estimate and the second baseline load estimate; The dividing the entire user cluster into a photovoltaic user cluster and a non-photovoltaic user cluster according to the generalized weather type label comprises: Extracting the average net load power of each user in the entire user cluster under different generalized weather type labels; For the entire user cluster, extract the first feature F1 and the second feature F2, and combine the first feature F1 and the second feature F2 to form a feature vector F n =[F1,F2] T ; According to the feature vector F n , dividing the entire user cluster into the photovoltaic user cluster and the non-photovoltaic user cluster; The expression of the first feature F1 is: Among them, t s and t e is the time sampling point, t s is the starting time point, t e is the end time point, represents the average net load power of user n at sampling point t under the D-th weather condition, represents the average net load power of user n at sampling point t under Class A weather conditions; The process of determining the second feature F2 is: According to the preset time window α=[t s ,t e ], there is a connection starting point t s and end point t e The sampling points y(t) of each time period on the straight line are expressed as: Where, t s and t e is the sampling point in the determined time window α, represents the average net load power of user n at sampling point t under the A-th weather condition. The sampling points y(t) in each time period on the straight line constitute a set S, S = {y(t)|t∈[t s ,t e ]}; The set of sampling points below the line S It is expressed as: The calculation formula of the second feature F2 is obtained as follows: F2=card( S ) / card(S) In the formula, card(·) represents the number of elements in the set.

2. The photovoltaic user baseline load estimation method according to claim 1, characterized in that: The step of clustering the observable photovoltaic output data to obtain a plurality of photovoltaic output power sets, setting generalized weather type labels for the plurality of photovoltaic output power sets, and obtaining generalized weather type labels for each day of the year includes: Extracting the observable photovoltaic output data of the part of photovoltaic users throughout the year to obtain a photovoltaic output curve set throughout the year; Clustering the photovoltaic output curve set throughout the year to obtain multiple photovoltaic output power sets, and calculating the average photovoltaic output of the multiple photovoltaic output power sets; each photovoltaic output power set contains multiple photovoltaic output curves; The photovoltaic output values ​​of the multiple photovoltaic output power sets at typical photovoltaic power generation moments are sorted, and weather labels of the multiple photovoltaic output power sets are set according to the sorting results to obtain generalized weather type labels for each day of the year.

3. The photovoltaic user baseline load estimation method according to claim 1, characterized in that: The step of establishing a support vector machine regression model according to the plurality of net load power sets, and determining a first baseline load estimation value of a photovoltaic user cluster according to the support vector machine regression model, comprises: Accumulating the net load power of each photovoltaic user in the photovoltaic user cluster to obtain the net load power of the photovoltaic user cluster; According to the generalized weather type label, the net load power of the photovoltaic user cluster is classified according to the label to obtain multiple net load power sets corresponding to the generalized weather type label; A support vector machine regression model is established for each of the plurality of net load powers, and a first load estimation value of the photovoltaic user cluster is determined according to the support vector machine regression model.

4. The photovoltaic user baseline load estimation method according to claim 1, characterized in that: The step of establishing a piecewise linear regression model based on the net load power and temperature data of the non-photovoltaic user cluster, and determining a second baseline load estimation value based on the piecewise linear regression model, includes: The first net load power of each non-photovoltaic user in the non-photovoltaic user cluster is accumulated to obtain the second net load power of the non-photovoltaic user cluster; the first net load power is the net load power of the non-photovoltaic user at the sampling point, and the second net load power is the net load power of the non-photovoltaic user cluster at the sampling point, and the sampling point is a pre-divided time node; Determining the net load power of the non-photovoltaic user cluster within a preset time period according to the second net load power; Segmenting outdoor temperature data samples at multiple sampling points within a preset time period to obtain segmented outdoor temperature sample data; According to the preset time period and the segmented outdoor temperature sample data, a segmented linear regression model corresponding to the preset time period is established; Obtaining the first outdoor temperature collected at each sampling point, where the first outdoor temperature is the outdoor temperature collected at each time sampling point in the target time period; The first outdoor temperature is segmented to obtain a second outdoor temperature, and the second outdoor temperature is input into the segmented linear regression model corresponding to the target time period to obtain the second baseline load estimation value.

5. A photovoltaic user baseline load estimation device, characterized in that: include: The information acquisition module is used to obtain the net load power and temperature data of all user clusters, as well as the observable photovoltaic output data of some photovoltaic users throughout the year; A classification module clusters the photovoltaic output data throughout the year to obtain multiple photovoltaic output power sets, sets generalized weather type labels for the multiple photovoltaic output power sets, and obtains generalized weather type labels for each day of the year; and divides the entire user cluster into a photovoltaic user cluster and a non-photovoltaic user cluster according to the generalized weather type labels; A first load estimation module is used to divide the net load power of the photovoltaic user cluster into multiple net load power sets according to the generalized weather type label, establish a support vector machine regression model according to the multiple net load power sets, and determine a first baseline load estimation value; A second load estimation module is used to establish a piecewise linear regression model according to the net load power and temperature data of the non-photovoltaic user cluster, and determine a second baseline load estimation value according to the piecewise linear regression model; A baseline load determination module, configured to determine a baseline load estimate of the entire user cluster based on the first baseline load estimate and the second baseline load estimate; In the classification module, the method of dividing the entire user cluster into a photovoltaic user cluster and a non-photovoltaic user cluster according to the generalized weather type label includes: Extracting the average net load power of each user in the entire user cluster under different generalized weather type labels; For the entire user cluster, extract the first feature F1 and the second feature F2, and combine the first feature F1 and the second feature F2 to form a feature vector F n =[F1,F2] T ; According to the feature vector F n , dividing the entire user cluster into the photovoltaic user cluster and the non-photovoltaic user cluster; The expression of the first feature F1 is: Among them, t s and t e is the time sampling point, t s is the starting time point, t e is the end time point, represents the average net load power of user n at sampling point t under the D-th weather condition, represents the average net load power of user n at sampling point t under Class A weather conditions; The process of determining the second feature F2 is: According to the preset time window α=[t s ,t e ], there is a connection starting point t s and end point t e The sampling points y(t) of each time period on the straight line are expressed as: Where, t s and t e is the sampling point in the determined time window α, represents the average net load power of user n at sampling point t under the A-th weather condition. The sampling points y(t) in each time period on the straight line constitute a set S, S = {y(t)|t∈[t s ,t e ]}; The set of sampling points below the line S It is expressed as: The calculation formula of the second feature F2 is obtained as follows: F2=card( S ) / card(S) In the formula, card(·) represents the number of elements in the set.

6. A terminal device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the computer program, the steps of the photovoltaic user baseline load estimation method according to any one of claims 1 to 4 are implemented.

7. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the steps of the photovoltaic user baseline load estimation method according to any one of claims 1 to 4 are implemented.

Images