Electric power material consumption prediction method and system based on cumulative difference, medium and processor

Abstract

The invention provides an electric power material consumption prediction method based on cumulative difference. The method comprises the following steps: S1, collecting historical data of electric power material consumption; s2, preprocessing the historical data to obtain a stable time sequence; s3, performing item-by-item cumulative calculation on the preprocessed historical data according to time steps to enable the data to be in an incremental trend so as to solve the characteristics of large data volatility and strong burstiness; s4, an SVR model is constructed and trained, and the trained SVR model is used to predict a newest cumulative value; and S5, carrying out difference on the latest cumulative sum value obtained by prediction and the last cumulative sum data in the historical data to obtain a consumption prediction value of the latest month. The method is devoted to fusion of the power industry and the data mining technology, the historical data characteristics of historical power material consumption are deeply analyzed, a preprocessing method and an accumulation method are adopted for processing according to the characteristic of large volatility, and accurate prediction of future consumption of power materials is achieved.

Description

Technical Field

[0001] The present invention relates to the technical field of material consumption prediction, and in particular to a method, system, medium and processor for predicting power material consumption based on cumulative difference. Background Art

[0002] The huge scale of my country's power grid provides solid support for the country's economic development and people's lives. However, this huge system also brings many challenges to the operation and maintenance management of enterprises, especially in the key area of ​​power material consumption forecasting. Insufficient forecasting may lead to material shortages and affect the normal operation of the power grid; while excessive forecasting may cause inventory backlogs and reduce the efficiency of corporate capital use. Therefore, achieving accurate forecasting of power material consumption is crucial to ensuring the stable operation of the power grid and improving the economic benefits of enterprises.

[0003] The consumption of electric power materials is usually recorded at monthly intervals, and the consumption in each month of the year varies greatly, showing huge volatility. Therefore, before conducting the analysis of the consumption of electric power materials, it is necessary to preprocess the historical data to obtain a relatively stable time series. There are three data stabilization methods currently used in material consumption forecasting methods: (1) Difference method, which eliminates trends and seasonality by subtracting the previous value, but for complex non-stationary series, simple differences may not be sufficient for complete stabilization. (2) Moving average method, which can smooth time series and reduce noise interference, but it is only suitable for stabilizing short-term fluctuations, and has poor effects on long-term trends and seasonal components. It also depends on the selection of the smoothing window, which may affect the results if it is too large or too small. (3) Exponential smoothing method, which is suitable for time series with trends and seasonal components, including simple exponential smoothing, Holt linear trend, Holt-Winters seasonal smoothing and other methods. It can adapt to different trends and seasonal changes, but for nonlinear or sudden data, the effect may not be ideal and also depends on the adjustment of model parameters. Therefore, the current method for predicting material consumption in power grid operation has poor compatibility in data processing and cannot cope well with long-term, non-stationary, and nonlinear data, resulting in unsatisfactory prediction results.

[0004] In view of this, a method, system, medium and processor for predicting power material consumption based on cumulative difference are needed. Summary of the invention

[0005] In view of the poor compatibility of data processing in the prior art, the inability to deal well with long-term, non-stationary, and nonlinear data, resulting in unsatisfactory prediction results, the present invention provides a method, system, medium, and processor for predicting power material consumption based on cumulative difference, which can perform pre-processing and cumulative calculation of data by deeply analyzing the characteristics of historical data of power material consumption, and realize accurate prediction of future power material consumption. The specific technical solution is as follows:

[0006] A power material consumption prediction method based on cumulative sum and difference, comprising:

[0007] S1: Collect historical data of power material consumption;

[0008] S2: Preprocess the historical data to obtain a stationary time series;

[0009] S3: Perform cumulative sum calculation item by item on the preprocessed historical data according to time steps to make the data show an increasing trend, so as to solve the problems of large data volatility and strong suddenness;

[0010] S4: Construct an SVR model and train it, and use the trained SVR model to predict the latest cumulative sum value;

[0011] S5: Take the difference between the predicted latest cumulative sum value and the last cumulative sum data in the historical data to obtain the consumption prediction value of the latest month.

[0012] Further, in step S2, the preprocessing of the historical data includes outlier correction and normalization processing; the outlier correction refers to using Z-score to detect outliers and replacing the outliers with moving average values, including the following steps:

[0013] S21: Organize the historical data into input data, and the formula is as follows:

[0014] {X} = {x1, x2,... x N};

[0015] S22: Solve the mean and standard deviation of the input data, and the formula is as follows:

[0016]

[0017] S23: Calculate the Z-score value of each data point in the input data, and the formula is as follows:

[0018]

[0019] S24: Determine outliers according to the size of Zi; if Zi > Tz, the deviation degree of the data point x i is very large, and it is determined as an outlier, otherwise Zi < Tz is determined as a normal point;

[0020] S25: For outliers, take the mean of k points on the left and right of it in time series to replace the outliers, that is:

[0021]

[0022] In the above formula, x it is xi The replacement value of x i is the ith data point; Tz is the outlier threshold; μ is the mean; σ is the standard deviation.

[0023] Furthermore, in step S2, the normalization process refers to scaling the time series data to a range of 0 to 1 to reduce the impact of data fluctuations; the data at the i-th step after normalization is x i ′:

[0024]

[0025] In the above formula, x max and x min Represents the maximum and minimum values ​​of the data in this set; x i is the i-th data point; x i ′ is the i-th data point after normalization.

[0026] Furthermore, the preprocessed historical data is cumulatively calculated item by item according to the time step, and the formula is as follows:

[0027] S={s1,s2,…,s N}={x′1,x′1+x′2,…,x′1+…+x′ N};

[0028] Among them, S is the accumulated data column.

[0029] Furthermore, the construction of the SVR model and training specifically include the following steps:

[0030] S41: Define the historical data sequence s1, s2, ..., s i-1 ;

[0031] S42: Determine the characteristics of the training sample; for the prediction s i , using the latest n samples s i-n ,s i-n+1 ,…,s i-1 and index i as feature; let feature vector x = (s i-n ,s i-n+1 ,…,s i-1 ,i);

[0032] S43: Construct a training set; for i=n+1,n+2..., there are training samples (x i-n ,s i ),in:

[0033] x i-n =(s i-n-1 ,s i-n ,…,s i-2 ,in);

[0034] S44: Build and train the SVR model; the goal is to find a function f(x) such that f(x i-n )≈s i ; The SVR model is trained by minimizing the loss function; the loss function is:

[0035]

[0036] In the above formula, λ is the regularization parameter and ω is the model parameter.

[0037] Further, the method of using the trained SVR model to predict the latest cumulative value includes the following steps:

[0038] Given the latest n samples s i-n ,s i-n+1 ,…,s i-1 And index i, forming the feature vector:

[0039] x=(s i-n ,s i-n+1 ,…,s i-1 ,i);

[0040] Use the objective function obtained during training to predict s i ,Right now:

[0041]

[0042] In the above formula, is the predicted value of the latest cumulative value; f(x) is the objective function obtained during training.

[0043] Furthermore, the consumption forecast value for the latest month is obtained by taking the difference between the latest cumulative value predicted and the last cumulative data in the historical data. The formula is as follows:

[0044] x′ N+1 =(x′1+…+x′ N +x′ N+1 )-(x′1+…+x′ N );

[0045] x′ N+1 It is the consumption forecast value for the latest month.

[0046] A power material consumption prediction system based on cumulative difference is applied to the power material consumption prediction method based on cumulative difference described above, comprising:

[0047] A collection module, which is used to collect historical data on power material consumption;

[0048] A processing module, which is used to pre-process historical data to obtain a stable time series;

[0049] The accumulation module is used to perform cumulative calculations on the preprocessed historical data item by item according to the time step, so that the data presents an increasing trend to solve the characteristics of large data volatility and strong suddenness;

[0050] The prediction module is used to build and train the SVR model and use the trained SVR model to predict the latest cumulative value;

[0051] The difference module is used to differentiate the latest cumulative value predicted from the last cumulative data in the historical data to obtain the consumption forecast value for the latest month.

[0052] A computer-readable storage medium includes a stored program, wherein when the program is executed, the device where the computer-readable storage medium is located is controlled to execute the above-mentioned method for predicting power material consumption based on cumulative difference.

[0053] A processor is used to run a program, wherein the program executes the above-mentioned power material consumption prediction method based on cumulative difference when running.

[0054] Compared with the prior art, the present invention has the following beneficial effects:

[0055] 1. We are committed to the integration of the power industry and data mining technology. Through in-depth analysis of the historical data characteristics of power material consumption, we use preprocessing methods and cumulative methods to process the data with large volatility, so as to achieve accurate prediction of future power material consumption.

[0056] 2. The implementation of this plan will effectively overcome the human errors that currently exist in the material demand forecasting of power companies, thereby ensuring that the safety and economy of power grid operations are significantly improved.

[0057] 3. Through scientific data-driven methods, not only the accuracy of prediction is improved, but also strong data support is provided for resource planning and management decisions of power companies. BRIEF DESCRIPTION OF THE DRAWINGS

[0058] In order to more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the following is a brief introduction to the drawings required for the specific embodiments or the description of the prior art. In all the drawings, similar elements or parts are generally identified by similar reference numerals. In the drawings, the elements or parts are not necessarily drawn according to the actual scale.

[0059] Figure 1 It is a flowchart of a method for predicting power material consumption based on cumulative difference;

[0060] Figure 2 This is a schematic diagram of the fluctuation of material consumption in a power supply station;

[0061] Figure 3 It is a schematic diagram of the increasing trend of the cumulative data column;

[0062] Figure 4 This is a schematic diagram of the latest predicted consumption values. DETAILED DESCRIPTION

[0063] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0064] It should be understood that when used in this specification and the appended claims, the terms "include" and "comprises" indicate the presence of described features, integers, steps, operations, elements and / or components, but do not exclude the presence or addition of one or more other features, integers, steps, operations, elements, components and / or combinations thereof.

[0065] It should also be understood that the terms used in the present specification are only for the purpose of describing specific embodiments and are not intended to limit the present invention. As used in the present specification and the appended claims, unless the context clearly indicates otherwise, the singular forms "a", "an" and "the" are intended to include plural forms.

[0066] It should be further understood that the term "and / or" used in the present description and the appended claims refers to any and all possible combinations of one or more of the associated listed items, and includes these combinations.

[0067] Embodiment 1

[0068] like Figure 1 The figure shows a flow chart of a method for predicting power material consumption based on cumulative difference, which includes the following steps:

[0069] S1: Collect historical data on power material consumption.

[0070] S2: Based on historical data, its consumption characteristics are characterized by large fluctuations and strong suddenness. In order to make the prediction results accurate, necessary data preprocessing is performed on the historical data to obtain a stable time series.

[0071] Furthermore, the necessary data preprocessing of the historical data includes outlier correction and normalization processing.

[0072] Further, the outlier correction refers to detecting outliers using Z-score and replacing the outliers with the moving average, including the following steps:

[0073] S21: Organize the historical data into input data:

[0074] {X} = {x1, x2,....x N};

[0075] S22: Solve the mean and standard deviation of the input data as:

[0076]

[0077]

[0078] S23: Calculate the Z-score value of each data point in the input data:

[0079]

[0080] S24: Determine the outliers according to the magnitude of Zi; if Zi > Tz, the deviation degree of the data point x i is very large, and it is determined as an outlier, otherwise Zi < Tz is determined as a normal point;

[0081] S25: For the outliers, take the k points on the left and right in its time series, calculate the mean value, and replace the outliers, that is

[0082]

[0083] In the above formula, x it is the replacement value of x i ; x i is the i-th data point; Tz is the outlier threshold; μ is the mean value; σ is the standard deviation.

[0084] Further, the normalization process refers to scaling the time series data to the range of 0 to 1 to reduce the influence of data fluctuations; the data at the i-th step of time after normalization is x i ':

[0085]

[0086] In the above formula, x max and x min represent the maximum and minimum values of the data in this set; x i is the i-th data point; x i ' is the i-th data point after normalization, and the data is converted to the [0, 1] interval after normalization.

[0087] Furthermore, the normalized data set is as follows:

[0088] D={x′1,x′2,…,x′ N}.

[0089] S3: The pre-processed historical data is cumulatively calculated item by item according to the time step, so that the data shows an increasing trend to solve the characteristics of large volatility and suddenness of power material data.

[0090] For example, the formula D = {x′1, x′2,…, x′ N} represents the time series data after outlier correction and normalization. Take the historical data of a real power supply station in a certain place as an example. Figure 2 shown.

[0091] Perform cumulative processing on the D data column, such as S = {s1, s2, ..., s N}={x′1,x′1+x′2,…,x′1+…+x′ N} is the accumulated data, such as Figure 3 As shown, the S data column shows an increasing trend.

[0092] S4: Build and train the SVR model, and use the trained SVR model to predict the latest cumulative value, which is x′1+…+x′ N +x′ N+1 .

[0093] Considering that the recorded data of power materials are generally small samples (generally recorded at monthly intervals starting from 2020), the SVR model suitable for small sample prediction tasks is used to perform x′1+…+x′ N +x′ N+1 The prediction is:

[0094]

[0095] Furthermore, the SVR model is constructed and trained, and the latest cumulative value is predicted using the trained SVR model, specifically including the following steps:

[0096] Extract n historical cumulative data from the S data column to predict the next cumulative data. Since the consumption of electric power materials varies greatly in different months, the index i of the month to be predicted is embedded in the input to facilitate model recognition rules.

[0097] S41: Define the historical data sequence s1, s2, ..., s i-1 ;

[0098] S42: Determine the characteristics of the training sample;

[0099] Specifically, for the prediction si , using the latest n samples s i-n ,s i-n+1 ,…,s i-1 and index i as feature;

[0100] Let the eigenvector x = (s i-n ,s i-n+1 ,…,s i-1 ,i);

[0101] S43: build training set;

[0102] For i=n+1,n+2..., there are training samples (x i-n ,s i ),in:

[0103] x i-n =(s i-n-1 ,s i-n ,…,s i-2 ,in);

[0104] S44: Build and train the SVR model;

[0105] The goal is to find a function f(x) such that f(x i-n )≈s i ;

[0106] By minimizing the loss function To train the SVR model;

[0107] In the above formula, λ is the regularization parameter and ω is the model parameter.

[0108] S45: Use the trained SVR model to predict the latest cumulative value;

[0109] Given the latest n samples s i-n ,s i-n+1 ,…,s i-1 And index i, forming the feature vector:

[0110] x=(s i-n ,s i-n+1 ,…,s i-1 ,i);

[0111] Use the objective function obtained during training to predict s i ,Right now:

[0112]

[0113] In the above formula, s i is the true value, For i The predicted value of .

[0114] S5: Difference the latest cumulative value obtained by prediction and the last cumulative data in the historical data to obtain the consumption prediction value of the latest month.

[0115] After getting the latest cumulative prediction value x′1+…+x′ N +x′ N+1 , we only need to differentiate it from the last cumulative data in the historical data to get: x′ N+1 =(x′1+…+x′ N +x′ N+1 )-(x′1+…+x′ N ), that is, to obtain the consumption forecast value of the latest month. Figure 4 As shown, the last point is the latest predicted consumption value.

[0116] This embodiment scheme is committed to the integration of the power industry and data mining technology. By deeply analyzing the historical data of historical power material consumption, it aims to achieve accurate prediction of future power material consumption. The implementation of this scheme will effectively overcome the human errors existing in the material demand forecasting of current power companies, thereby ensuring that the safety and economy of power grid operation are significantly improved. Through scientific data-driven methods, this scheme not only improves the accuracy of prediction, but also provides strong data support for resource planning and management decisions of power companies.

[0117] Embodiment 2

[0118] A power material consumption prediction system based on cumulative difference is applied to the power material consumption prediction method based on cumulative difference described above, comprising:

[0119] A collection module, which is used to collect historical data on power material consumption;

[0120] A processing module, which is used to pre-process historical data to obtain a stable time series;

[0121] The accumulation module is used to perform cumulative calculations on the preprocessed historical data item by item according to the time step, so that the data presents an increasing trend to solve the characteristics of large data volatility and strong suddenness;

[0122] The prediction module is used to build and train the SVR model and use the trained SVR model to predict the latest cumulative value;

[0123] The difference module is used to differentiate the latest cumulative value predicted from the last cumulative data in the historical data to obtain the consumption forecast value for the latest month.

[0124] Embodiment 3

[0125] A computer-readable storage medium includes a stored program, wherein when the program is executed, the device where the computer-readable storage medium is located is controlled to execute the above-mentioned method for predicting power material consumption based on cumulative difference.

[0126] Embodiment 4

[0127] A processor is used to run a program, wherein the program executes the above-mentioned power material consumption prediction method based on cumulative difference when running.

[0128] The present application provides a method for predicting power material consumption based on cumulative difference, including: S1: collecting historical data of power material consumption; S2: preprocessing historical data to obtain a stable time series; S3: performing cumulative calculations on the preprocessed historical data item by item according to time steps, so that the data presents an increasing trend, in order to solve the characteristics of large data volatility and strong suddenness; S4: constructing and training an SVR model, and using the trained SVR model to predict the latest cumulative value; S5: performing differential calculations on the predicted latest cumulative value and the last cumulative data in the historical data to obtain the consumption forecast value of the latest month. The application is committed to the integration of the power industry and data mining technology, and through in-depth analysis of the characteristics of historical data of power material consumption, preprocessing methods and cumulative methods are used to process the data according to the characteristics of large volatility, so as to achieve accurate prediction of future power material consumption.

[0129] Those of ordinary skill in the art will appreciate that the units of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of the two. In order to clearly illustrate the interchangeability of hardware and software, the composition of each example has been generally described in terms of function in the above description. 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.

[0130] In the embodiments provided by the present invention, it should be understood that the division of units is only a logical function division, and there may be other division methods in actual implementation, for example, multiple units can be combined into one unit, one unit can be split into multiple units, or some features can be ignored, etc.

[0131] 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.

[0132] If the integrated 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 technical solution of the present invention, in essence, or the part that contributes to the prior art, or all or part of the technical solution can be embodied in the form of a software product, which is stored in a storage medium and includes several instructions for a computer device (which can be a personal computer, a server or a network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes: U disk, read-only memory (ROM, Read-0nlyMemory), random access memory (RAM, RandomAccessMemory), mobile hard disk, magnetic disk or optical disk, etc., which can store program code.

[0133] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or replace some or all of the technical features therein by equivalents. These modifications or replacements do not make the essence of the corresponding technical solutions deviate from the scope of the technical solutions of the embodiments of the present invention, and they should all be included in the scope of the claims and specification of the present invention.

Claims

1. A method for predicting power material consumption based on cumulative difference, It is characterized in that include: S1: Collect historical data on power material consumption; S2: Preprocess the historical data to obtain a stable time series; S3: The pre-processed historical data is cumulatively calculated item by item according to the time step, so that the data presents an increasing trend to solve the characteristics of large data volatility and strong suddenness; S4: Build and train the SVR model, and use the trained SVR model to predict the latest cumulative value; S5: Difference the latest cumulative value obtained by prediction and the last cumulative data in the historical data to obtain the consumption prediction value of the latest month.

2. According to the method for predicting power material consumption based on cumulative difference according to claim 1, It is characterized in that In step S2, the preprocessing of historical data includes outlier correction and normalization processing; the outlier correction refers to using Z-score to detect outliers and using sliding average to replace outliers, including the following steps: S21: Organize historical data into input data, the formula is as follows: {X}={x 1 ,x 2 ,....x N }; S22: Solve the mean and standard deviation of the input data. The formula is as follows: S23: Calculate the Z-score value of each data point in the input data. The formula is as follows: S24: Determine outliers based on the magnitude of Zi; if Zi > Tz, the deviation degree of the data point x i is very large, and it is determined as an outlier; otherwise, if Zi < Tz, it is determined as a normal point; S25: For outliers, use the k points on the left and right of the time series to find the mean and replace the outliers, that is: In the above formula, x it For x i The replacement value of x i is the ith data point; Tz is the outlier threshold; μ is the mean; σ is the standard deviation.

3. According to claim 2, the power material consumption prediction method based on cumulative difference, It is characterized in that In step S2, the normalization process refers to scaling the time series data to a range of 0 to 1 to reduce the impact of data fluctuations; the data of the i-th step after normalization is x i ′: In the above formula, x max and x min Represents the maximum and minimum values ​​of the data in this set; x i is the i-th data point; x i ′ is the i-th data point after normalization.

4. The power material consumption prediction method based on cumulative difference according to claim 3, It is characterized in that The preprocessed historical data is cumulatively calculated item by item according to the time step, and the formula is as follows: S={s 1 ,s 2 ,…,s N }={x 1 ′,x 1 ′+x′ 2 ,…,x 1 ′+…+x′ N }; Among them, S is the accumulated data column.

5. According to claim 4, the method for predicting power material consumption based on cumulative difference, It is characterized in that The SVR model is constructed and trained, specifically including the following steps: S41: Define historical data sequence s 1 ,s 2 ,…,s i-1 ; S42: Determine the characteristics of the training sample; for the prediction s i , using the latest n samples s i-n ,s i-n+1 ,…,s i-1 and index i as feature; let feature vector x = (s i-n ,s i-n+1 ,…,s i-1 ,i); S43: Construct a training set; for i=n+1,n+2..., there are training samples (x i-n ,s i ),in: x i-n =(s i-n-1 ,s i-n ,…,s i-2 ,i-n); S44: Build and train the SVR model; the goal is to find a function f(x) such that f(x i-n )≈s i ; The SVR model is trained by minimizing the loss function; the loss function is: In the above formula, λ is the regularization parameter and ω is the model parameter.

6. The method for predicting power material consumption based on cumulative difference according to claim 5, It is characterized in that The method of using the trained SVR model to predict the latest cumulative value includes the following steps: Given the latest n samples s i-n ,s i-n+1 ,…,s i-1 And index i, forming the feature vector: x=(s i-n ,s i-n+1 ,…,s i-1 ,i); Use the objective function obtained during training to predict s i ,Right now: In the above formula, is the predicted value of the latest cumulative value; f(x) is the objective function obtained during training.

7. The method for predicting power material consumption based on cumulative difference according to claim 4, It is characterized in that The consumption forecast value for the latest month is obtained by taking the difference between the latest cumulative value predicted and the last cumulative data in the historical data. The formula is as follows: x′ N+1 =(x 1 ′+…+x′ N +x′ N+1 )-(x 1 ′+…+x′ N ); x′ N+1 It is the consumption forecast value for the latest month.

8. A power material consumption prediction system based on cumulative difference, It is characterized in that The power material consumption prediction method based on cumulative difference applied to any one of claims 1 to 7 comprises: A collection module, which is used to collect historical data on power material consumption; A processing module, which is used to pre-process historical data to obtain a stable time series; The accumulation module is used to perform cumulative calculations on the preprocessed historical data item by item according to the time step, so that the data presents an increasing trend to solve the characteristics of large data volatility and strong suddenness; The prediction module is used to build and train the SVR model and use the trained SVR model to predict the latest cumulative value; The difference module is used to differentiate the latest cumulative value predicted from the last cumulative data in the historical data to obtain the consumption forecast value for the latest month.

9. A computer-readable storage medium, It is characterized in that The computer-readable storage medium includes a stored program, wherein when the program is executed, the device where the computer-readable storage medium is located is controlled to execute the power material consumption prediction method based on cumulative difference as described in any one of claims 1 to 7.

10. A processor, It is characterized in that The processor is used to run a program, wherein the program, when running, executes the power material consumption prediction method based on cumulative difference as described in any one of claims 1 to 7.

Images