New energy power prediction correction method and system based on mixed channel cloud model

Abstract

The invention discloses a new energy power prediction correction method and system based on a mixed channel cloud model, and the method comprises the steps: obtaining historical key data, including a predicted power value and an actual power value, of new energy power prediction from different data channels, and obtaining the predicted power value and the actual power value after normalization processing through normalization processing; constructing a cloud function by using a triangular membership function based on the predicted power value and the actual power value after normalization processing; and constructing a mixed channel cloud model, generating a correction weight for the power prediction value of each time point in the target time period by a cloud function, performing automatic dynamic component optimization on the power prediction value of each time point, correcting the power prediction value of the corresponding time point, and obtaining a final new energy prediction power value. According to the method, the power prediction value can be dynamically adjusted and optimized to correct the error in prediction, and the precision, reliability and stability of the prediction result of new energy such as wind energy and solar energy can be remarkably improved.

Description

Technical Field

[0001] The present invention belongs to the technical field of renewable energy power generation, and relates to a renewable energy power prediction and correction method and system based on a hybrid channel cloud model. Background Art

[0002] In the context of the current energy structure transformation, renewable energy sources such as solar energy and wind energy are widely used due to their clean and renewable characteristics. However, the power output of these renewable energy sources is strongly affected by climate and weather conditions and exhibits high volatility, which poses new challenges to the stable operation and management of the power system.

[0003] In order to ensure the balance of electricity supply and demand and their real-time matching, accurate power forecasting has become an indispensable technology in power grid operation, which can help dispatchers make effective power generation plans to adapt to changing electricity demand.

[0004] Traditional forecasting methods mostly use the correlation between meteorological factors such as wind speed, temperature, humidity and renewable energy power generation for forecasting. However, there is a complex correlation between these meteorological factors and renewable energy power generation. These variables are not only variable in time, but also show obvious diversity and periodicity in spatial distribution. This diversity and periodicity make it difficult for the forecasting model to fully capture and accurately describe the dynamic relationship between these variables, so the results of the traditional forecasting model need to be further optimized and corrected. Summary of the invention

[0005] In order to address the deficiencies in the prior art, the present invention provides a new energy power prediction and correction method and system based on a hybrid channel cloud model, which analyzes historical key data, including predicted power values ​​and actual power values, and corrects the predicted values ​​based on the proposed cloud model hybrid channel and automatic dynamic component optimization. This can significantly improve the prediction capability of the new energy power system, effectively deal with the uncertainty and volatility of power output, and provide strong support for the stable operation of the power grid.

[0006] The present invention adopts the following technical solution.

[0007] A first aspect of the present invention provides a new energy power prediction and correction method based on a hybrid channel cloud model, comprising:

[0008] Obtain historical key data of new energy power prediction from different data channels, including predicted power value and actual power value, and obtain normalized predicted power value and actual power value through normalization processing;

[0009] Based on the normalized predicted power value and actual power value, the cloud function is constructed using the triangular membership function;

[0010] A hybrid channel cloud model is constructed, and the cloud function generates correction weights for the power prediction value at each time point in the target time period, so as to automatically optimize the power prediction value at each time point dynamically, correct the power prediction value at the corresponding time point, and obtain the final new energy prediction power value.

[0011] Preferably, the cloud function is constructed by using a triangular membership function based on the predicted power value and the actual power value after normalization, including:

[0012] According to the predicted power value P' after normalization predicted And the actual power value P' actual , calculate the key parameters of the corresponding triangular membership function;

[0013] Based on the key parameters, the normalized predicted power value P' is calculated using the triangular membership function. predicted Or the actual power value P' actual After processing, the cloud membership value M(x′; μ, a, b) is obtained;

[0014] Using the cloud membership value M of all data channels i (x′; μ i ,a i ,b i )’s mean μ1 and standard deviation σ1, construct cloud functions f(x′; μ1, σ1) of different data channels;

[0015] Preferably, the key parameters include peak μ, left edge a and right boundary b.

[0016] Preferably, the calculation formula of the peak value μ is:

[0017] μ=mean(x′) (2)

[0018] The left border a The calculation formula is:

[0019] a=μ-std(x′) (3)

[0020] The calculation formula of the right boundary b is:

[0021] b=μ+std(x′) (4)

[0022] Among them, mean() and std() are the mean function and standard deviation function of the corresponding data set respectively;

[0023] x′ is the predicted power value P' after normalization predicted Or the actual power value P' actual .

[0024] Preferably, the formula of the triangular membership function is:

[0025]

[0026] Among them, x′ is the predicted power value P' after normalization predicted Or the actual power value P' actual .

[0027] Preferably, the cloud functions of the different data channels are:

[0028]

[0029] Preferably, the hybrid channel cloud model is as follows:

[0030] P predicted.combined.i =w i P predicted.i (9)

[0031] Among them, w i is the power prediction value P at the i-th time point predicted.i The correction weight is obtained by processing the normalized predicted power value of the ith time point by the cloud function f(x′; μ1, σ1);

[0032] P predicted.combined.i It is the corrected power prediction value output at the i-th time point.

[0033] A second aspect of the present invention provides a new energy power prediction and correction system based on a hybrid channel cloud model, comprising:

[0034] A data acquisition module is used to obtain historical key data of new energy power prediction from different data channels, including predicted power values ​​and actual power values, and obtain normalized predicted power values ​​and actual power values ​​through normalization processing;

[0035] A function building module, used to build a cloud function using a triangular membership function based on the normalized predicted power value and the actual power value;

[0036] The power prediction correction module is used to build a hybrid channel cloud model. The cloud function generates correction weights for the power prediction value at each time point in the target time period, automatically optimizes the power prediction value at each time point, corrects the power prediction value at the corresponding time point, and obtains the final new energy prediction power value.

[0037] Preferably, the hybrid channel cloud model is as follows:

[0038] P predicted.combined.i =w i P predicted.i (9)

[0039] Among them, w i is the power prediction value P at the i-th time point predicted.i The correction weight is obtained by processing the normalized predicted power value of the ith time point by the cloud function f(x′; μ1, σ1);

[0040] P predicted.combined.i It is the corrected power prediction value output at the i-th time point.

[0041] Preferably, the cloud function f(x′; μ1, σ1) is:

[0042]

[0043] Where μ1 and σ1 are the mean and standard deviation of the cloud membership values ​​of all data channels;

[0044] x′ is the predicted power value after normalization.

[0045] A terminal comprises a processor and a storage medium; the storage medium is used to store instructions; the processor is used to operate according to the instructions to execute the steps of the method.

[0046] A computer-readable storage medium stores a computer program, which implements the steps of the method when executed by a processor.

[0047] Compared with the prior art, the beneficial effects of the present invention include at least:

[0048] The present invention designs a triangular membership function, processes the predicted power value and the actual power value after normalization, and obtains the cloud membership value, which lays the foundation for the calculation of key parameters of the cloud function. Compared with directly using the actual power value and the predicted power value to calculate the mean and standard deviation, some bad data can be filtered out and the parameter accuracy can be improved.

[0049] The present invention integrates power data of key dimensions from multiple sources and constructs a hybrid channel cloud model that takes the time dimension into consideration. It can integrate the information of each data channel and utilize the unique advantages and complementary characteristics of each channel to form a more comprehensive and accurate power prediction and correction model. It not only improves the prediction accuracy, but also enhances its adaptability to the power fluctuations and uncertainties of new energy, providing more reliable, efficient and stable support for the management and scheduling of new energy.

[0050] The hybrid channel cloud model of the present invention utilizes cloud functions and automatic dynamic component optimization strategies, which can dynamically adjust and optimize power prediction values ​​to correct errors in the prediction, solve the uncertainty problem in wind and solar power generation power prediction, and can significantly improve the accuracy, reliability and stability of prediction results of new energy such as wind energy and solar energy, providing powerful data support and intelligent decision-making tools for the reliable operation of new energy power systems. BRIEF DESCRIPTION OF THE DRAWINGS

[0051] Figure 1 Modeling flow chart for cloud model hybrid channel;

[0052] Figure 2 is the actual power diagram;

[0053] Figure 3 For the predicted power diagram;

[0054] Figure 4 is the corrected predicted power diagram;

[0055] Figure 5 It is the predicted power error curve diagram after prediction and correction. DETAILED DESCRIPTION

[0056] In order to make the purpose, technical scheme and advantages of the present invention clearer, the technical scheme of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. The embodiments described in this application are only part of the embodiments of the present invention, not all of them. Based on the spirit of the present invention, other embodiments obtained by ordinary technicians in this field without creative work are all within the scope of protection of the present invention.

[0057] like Figure 1 As shown, Embodiment 1 of the present invention provides a new energy power prediction correction method based on a hybrid channel cloud model, which corrects the power prediction data to improve the accuracy of wind and solar power generation power prediction, including data normalization processing, cloud function construction, data channel cloud model construction, etc., as follows:

[0058] Step 1: Obtain historical key data of new energy power forecast from different data channels, including the predicted power value P predicted And the actual power value P actual , and predict the power value P predicted And the actual power value P actual Perform normalization to obtain the predicted power value P' after normalization predicted And the actual power value P' actual ;

[0059] Further preferably, key data dimensions are collected, including the predicted power value P of the target time and spatial location predicted, actual power value P actual .

[0060] The collected predicted power value P predicted , actual power value P actual Normalization is performed, and the formula is:

[0061]

[0062] in x is the original data, and is the predicted power value P before normalization. predicted Or the actual power value P actual ; x' is the normalized data, which is the predicted power value P' after normalization predicted Or the actual power value P' actual .

[0063] Step 2: Based on the predicted power value P' after normalization predicted And the actual power value P' actual , using triangular membership functions to build cloud functions, including:

[0064] Step 201: Determine the key parameters of the appropriate triangular membership function for cloud model construction, including peak, left boundary and right boundary:

[0065] According to the predicted power value P' after normalization in step 1 predicted And the actual power value P' actual , determine the key parameters of the corresponding triangular membership function, including the peak μ, the left boundary a and right boundary b;

[0066] The peak value is denoted as μ, which is set to the average value of the data set and is calculated as:

[0067] μ=mean(x′) (2)

[0068] The left boundary is a , set to the peak minus the standard deviation, calculated as:

[0069] a=μ-std(x′) (3)

[0070] The right boundary is denoted as b, which is set to the peak value plus the standard deviation, and the calculation formula is:

[0071] b=μ+std(x′) (4)

[0072] Among them, mean() and std() are the mean function and standard deviation function of the data set respectively;

[0073] x′ is the predicted power value P' after normalization predictedOr the actual power value P' actual .

[0074] Step 202: Based on the key parameters, use the triangular membership function to normalize the predicted power value P' predicted Or the actual power value P' actual After processing, the cloud membership value M(x′; μ, a, b) is obtained;

[0075] Among them, the triangular membership function is used to construct the cloud membership of wind and solar uncertainty. The triangular membership function is used to describe the functional relationship between the input variable x and the cloud model peak μ, the left boundary a, and the right boundary b to form an accurate description of the cloud membership. The formula of the triangular membership function is:

[0076]

[0077] Among them, x′ is the predicted power value P' after normalization predicted Or the actual power value P' actual ;

[0078] When x′ is P' predicted When P' predicted Calculate the peak μ and left boundary a and right boundary b;

[0079] When x′ is P' actual When P' actual Calculate the peak μ and left boundary a and right boundary b.

[0080] Step 203: Using the cloud membership values ​​M of all data channels i (x′; μ i ,a i ,b i )’s mean μ1 and standard deviation σ1, construct cloud functions f(x′; μ1, σ1) of different data channels;

[0081] The cloud membership values ​​obtained in step 202 are used to calculate the mean and standard deviation to determine the key parameters of the cloud model and construct cloud functions for different data channels. The mean value, one of the key parameters of the cloud function, is denoted as μ1, which is the arithmetic mean of the triangular membership values ​​of all data points. The calculation formula is:

[0082]

[0083] One of the key parameters of the cloud function is the standard deviation, denoted as σ1, which is the square root of the average of the squares of the differences between each membership value and the mean. The calculation formula is:

[0084]

[0085] After determining the key parameters of the cloud function, use the cloud function to further process these membership values ​​to simulate the behavior and characteristics of the data.

[0086] The cloud function expression is:

[0087]

[0088] It can be understood that in the triangular membership function formula (5) corresponding to step 202, x′ is the predicted power value P′ after normalization processing. predicted Or the actual power value P' actual ; When x′ takes P' predicted When x′ takes P′ actual , and obtain the cloud membership value M2 of another data channel; therefore, step 202 only obtains the cloud membership values ​​of two data channels, namely M1 and M2;

[0089] Step 203 calculates the mean and standard deviation of M1 and M2 according to formulas (6)-(7) (n=2 in formulas (6)-(7)), obtains μ1 and σ1, and substitutes them into formula (8) to construct a cloud function f(x′; μ1, σ1) of different data channels, where x′ is the predicted power value P′ after normalization. predicted Or the actual power value P' actual .

[0090] The predicted power value P' at each time point after normalization is further processed using the cloud function f(x'; μ1, σ1) predicted.i , we can get the modified weight w of the predicted power value i .

[0091] Step 3: Build a hybrid channel cloud model to correct the predicted value of new energy power.

[0092] Construct a hybrid channel cloud model, and use the cloud function to generate correction weights for the power prediction value at each time point, form dynamic adjustment weights for the power prediction value at each time point within the target time period, automatically optimize the power prediction value dynamically, and correct the predicted power value of new energy.

[0093] P predicted.combined.i =w i P predicted.i (9)

[0094] Among them, w i is the power prediction value P at the i-th time point predicted.i The correction weight is obtained by processing the normalized predicted power value of the ith time point by the cloud function f(x′; μ1, σ1), that is, the normalized Ppredicted.i Substituting x′ into formula (9), we get w i ;

[0095] P predicted.combined.i It is the corrected power prediction value output at the i-th time point.

[0096] In order to further verify the feasibility of the above cloud model hybrid channel modeling process, the existing wind power data is used to simulate the cloud model hybrid channel modeling in MATLAB according to the process. The simulation data is as follows: Figure 2-4 As shown, the actual power, the predicted power, and the result diagram after the predicted power is corrected are displayed.

[0097] In order to further demonstrate the correction effect, the predicted power error curves after prediction and correction are given, as shown in Figure 5 As shown in the figure, the simulation results help to intuitively see that the corrected predicted power output by the cloud model hybrid channel is more accurate than the original predicted power, thus providing technical support for improving the subsequent power prediction accuracy.

[0098] Embodiment 2 of the present invention provides a new energy power prediction and correction system based on a hybrid channel cloud model, comprising:

[0099] A data acquisition module is used to obtain historical key data of new energy power prediction from different data channels, including predicted power values ​​and actual power values, and obtain normalized predicted power values ​​and actual power values ​​through normalization processing;

[0100] A function building module, used to build a cloud function using a triangular membership function based on the normalized predicted power value and the actual power value;

[0101] The power prediction correction module is used to build a hybrid channel cloud model. The cloud function generates correction weights for the power prediction value at each time point in the target time period, automatically optimizes the power prediction value at each time point, corrects the power prediction value at the corresponding time point, and obtains the final new energy prediction power value.

[0102] Embodiment 3 of the present invention provides a terminal, including a processor and a storage medium; the storage medium is used to store instructions; the processor is used to operate according to the instructions to execute the steps of the method.

[0103] Embodiment 4 of the present invention provides a computer-readable storage medium having a computer program stored thereon, which implements the steps of the method when executed by a processor.

[0104] Compared with the prior art, the beneficial effects of the present invention include at least:

[0105] The present invention designs a triangular membership function, processes the predicted power value and the actual power value after normalization, and obtains the cloud membership value, which lays the foundation for the calculation of key parameters of the cloud function. Compared with directly using the actual power value and the predicted power value to calculate the mean and standard deviation, some bad data can be filtered out and the parameter accuracy can be improved.

[0106] The present invention integrates power data of key dimensions from multiple sources and constructs a hybrid channel cloud model that takes the time dimension into consideration. It can integrate the information of each data channel and utilize the unique advantages and complementary characteristics of each channel to form a more comprehensive and accurate power prediction model. This not only improves the prediction accuracy of the model, but also enhances its adaptability to fluctuations and uncertainties in new energy power, providing more reliable, efficient and stable support for the management and scheduling of new energy.

[0107] The hybrid channel cloud model of the present invention utilizes cloud functions and automatic dynamic component optimization strategies, which can not only identify and simulate the periodic laws of meteorological and power data, but also dynamically adjust and optimize the power prediction model to correct errors in the prediction, solving the uncertainty problem in wind and solar power generation power prediction, and can significantly improve the accuracy, reliability and stability of the prediction results of new energy such as wind energy and solar energy, providing powerful data support and intelligent decision-making tools for the reliable operation of new energy power systems.

[0108] The present disclosure may be a system, a method and / or a computer program product. The computer program product may include a computer-readable storage medium carrying computer-readable program instructions for causing a processor to implement various aspects of the present disclosure.

[0109] A computer-readable storage medium may be a tangible device that can hold and store instructions used by an instruction execution device. A computer-readable storage medium may be, for example, but not limited to, an electrical storage device, a magnetic storage device, an optical storage device, an electromagnetic storage device, a semiconductor storage device, or any suitable combination of the foregoing. More specific examples of computer-readable storage media (a non-exhaustive list) include: a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), a static random access memory (SRAM), a portable compact disk read-only memory (CD-ROM), a digital versatile disk (DVD), a memory stick, a floppy disk, a mechanical encoding device, such as a punch card or a raised structure in a groove on which instructions are stored, and any suitable combination of the foregoing. As used herein, a computer-readable storage medium is not to be interpreted as a transient signal per se, such as a radio wave or other freely propagating electromagnetic wave, an electromagnetic wave propagating through a waveguide or other transmission medium (e.g., a light pulse through a fiber optic cable), or an electrical signal transmitted through a wire.

[0110] The computer-readable program instructions described herein can be downloaded from a computer-readable storage medium to each computing / processing device, or downloaded to an external computer or external storage device via a network, such as the Internet, a local area network, a wide area network, and / or a wireless network. The network can include copper transmission cables, optical fiber transmissions, wireless transmissions, routers, firewalls, switches, gateway computers, and / or edge servers. The network adapter card or network interface in each computing / processing device receives the computer-readable program instructions from the network and forwards the computer-readable program instructions for storage in the computer-readable storage medium in each computing / processing device.

[0111] The computer program instructions for performing the operation of the present disclosure may be assembly instructions, instruction set architecture (ISA) instructions, machine instructions, machine-related instructions, microcode, firmware instructions, state setting data, or source code or object code written in any combination of one or more programming languages, including object-oriented programming languages, such as Smalltalk, C++, etc., and conventional procedural programming languages, such as "C" language or similar programming languages. Computer-readable program instructions may be executed completely on a user's computer, partially on a user's computer, as an independent software package, partially on a user's computer, partially on a remote computer, or completely on a remote computer or server. In the case of a remote computer, the remote computer may be connected to the user's computer via any type of network, including a local area network (LAN) or a wide area network (WAN), or may be connected to an external computer (e.g., using an Internet service provider to connect via the Internet). In some embodiments, an electronic circuit, such as a programmable logic circuit, a field programmable gate array (FPGA), or a programmable logic array (PLA), may be customized by utilizing the state information of the computer-readable program instructions, and the electronic circuit may execute the computer-readable program instructions, thereby realizing various aspects of the present disclosure.

[0112] 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 above embodiments, ordinary technicians in the relevant field should understand that the specific implementation methods of the present invention can still be modified or replaced by equivalents, and any modifications or equivalent replacements that do not depart from the spirit and scope of the present invention should be covered within the scope of protection of the claims of the present invention.

Claims

1. A new energy power prediction and correction method based on a hybrid channel cloud model, characterized in that: include: Obtain historical key data of new energy power prediction from different data channels, including predicted power values ​​and actual power values, and obtain normalized predicted power values ​​and actual power values ​​through normalization processing; Based on the normalized predicted power value and actual power value, the cloud function is constructed using the triangular membership function; A hybrid channel cloud model is constructed, and the cloud function generates correction weights for the power prediction value at each time point in the target time period, so as to automatically optimize the power prediction value at each time point dynamically, correct the power prediction value at the corresponding time point, and obtain the final new energy prediction power value.

2. The new energy power prediction and correction method based on the hybrid channel cloud model according to claim 1 is characterized by: The cloud function is constructed based on the predicted power value and the actual power value after normalization using a triangular membership function, including: According to the predicted power value P' after normalization predicted And the actual power value P' actual , calculate the key parameters of the corresponding triangular membership function; Based on the key parameters, the normalized predicted power value P' is calculated using the triangular membership function. predicted Or the actual power value P' actual After processing, the cloud membership value M(x′; μ, a, b) is obtained; Using the cloud membership value M of all data channels i (x′; μ i ,a i ,b i )’s mean μ1 and standard deviation σ1, and construct cloud functions f(x′; μ1, σ1) of different data channels.

3. The new energy power prediction and correction method based on the hybrid channel cloud model according to claim 2 is characterized in that: The key parameters include peak μ, left boundary a and right boundary b.

4. The new energy power prediction and correction method based on the hybrid channel cloud model according to claim 3 is characterized by: The calculation formula of the peak value μ is: μ=mean(x′) (2) The calculation formula of the left boundary a is: a=μ-std(x′) (3) The calculation formula of the right boundary b is: b=μ+std(x′) (4) Among them, mean() and std() are the mean function and standard deviation function of the corresponding data set respectively; x′ is the predicted power value P' after normalization predicted Or the actual power value P' actual .

5. The new energy power prediction and correction method based on the hybrid channel cloud model according to claim 3 is characterized by: The formula of the triangular membership function is: Among them, x′ is the predicted power value P' after normalization predicted Or the actual power value P' actual .

6. The new energy power prediction and correction method based on the hybrid channel cloud model according to claim 2 is characterized by: The cloud functions of the different data channels are:

7. The new energy power prediction and correction method based on the hybrid channel cloud model according to claim 1 is characterized by: The hybrid channel cloud model is as follows: P predicted.combined.i =w i P predicted.i (9) Among them, w i is the power prediction value P at the i-th time point predicted.i The correction weight is obtained by processing the normalized predicted power value of the ith time point by the cloud function f(x′; μ1, σ1); P predicted.combined.i It is the corrected power prediction value output at the i-th time point.

8. A new energy power prediction and correction system based on a hybrid channel cloud model, characterized in that: The system comprises: A data acquisition module is used to obtain historical key data of new energy power prediction from different data channels, including predicted power values ​​and actual power values, and obtain normalized predicted power values ​​and actual power values ​​through normalization processing; A function building module, used to build a cloud function using a triangular membership function based on the normalized predicted power value and the actual power value; The power prediction correction module is used to build a hybrid channel cloud model. The cloud function generates correction weights for the power prediction value at each time point in the target time period, automatically optimizes the power prediction value at each time point, corrects the power prediction value at the corresponding time point, and obtains the final new energy prediction power value.

9. The new energy power prediction and correction system based on the hybrid channel cloud model according to claim 8 is characterized by: The hybrid channel cloud model is as follows: P predicted.combined.i =w i P predicted.i (9) Among them, w i is the power prediction value P at the i-th time point predicted.i The correction weight of is obtained by processing the normalized predicted power value of the ith time point by the cloud function f(x′; μ1, σ1); P predicted.combined.i It is the corrected power prediction value output at the i-th time point.

10. The new energy power prediction and correction system based on the hybrid channel cloud model according to claim 9 is characterized in that: The cloud function f(x′; μ1, σ1) is: Where μ1 and σ1 are the mean and standard deviation of the cloud membership values ​​of all data channels; x′ is the predicted power value after normalization.

11. A terminal comprising a processor and a storage medium; characterized in that: The storage medium is used to store instructions; The processor is configured to operate according to the instructions to execute the steps of the method according to any one of claims 1-7.

12. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the program is executed by a processor, the steps of the method according to any one of claims 1 to 7 are implemented.

Images