Ultra-short-term photovoltaic power range prediction methods, systems, devices and equipment

Abstract

This invention discloses a method, system, device, and equipment for ultra-short-term photovoltaic (PV) power range prediction. The method includes: acquiring PV power plant operation data and meteorological data; inputting the PV power plant operation data and meteorological data into a trained spatiotemporal feedforward neural network model to obtain a first single-value result; inputting the first single-value result into a probability distribution function, and calculating a second single-value result based on the probability distribution function; and simultaneously solving a system of equations combining the first and second single-value results, and outputting the PV power range prediction result based on the system of equations. This invention solves the problem of the effectiveness of PV power plant power prediction, improves the security of PV integration into the power system, and, moreover, allows PV power plants to increase their competitive advantage in electricity market decision-making.

Description

Technical Field

[0001] This invention relates to the fields of planning and artificial intelligence, new energy and photovoltaic power prediction technology, and in particular to ultra-short-term photovoltaic power range prediction methods, systems, devices and equipment. Background Technology

[0002] With the adjustment of the world's energy structure, the rational development and effective utilization of renewable energy have become increasingly important in the energy development strategies of various countries. Among them, solar energy is a clean energy source that is widely distributed, green, environmentally friendly, and easy to utilize. Solar power generation, as an important clean alternative and electricity substitute, is playing an increasingly important role in the energy supply system.

[0003] As the scale of photovoltaic (PV) installations continues to grow, the difficulty of grid integration is increasing, and the problem of curtailment is becoming increasingly prominent. Improving existing PV power prediction systems to enhance prediction accuracy and reliability, and providing more accurate prediction results to power system dispatch departments, will be key to solving the PV grid integration problem.

[0004] Photovoltaic power forecasting can be categorized by time frame into ultra-short-term, short-term, medium-term, and long-term forecasts. Ultra-short-term forecasts predict photovoltaic power output for the next 0-4 hours, and the results are generally used for power system peak shaving and frequency regulation, economic load dispatching, and spinning reserve adjustment. Short-term forecasts predict photovoltaic power output for the next 1-3 days, and the results are generally used for planning power generation one day in advance, optimizing cold and hot reserves, and dispatching grid resources. Medium-term forecasts predict photovoltaic power output for the next few weeks to several months, and the results are typically used for scheduling maintenance at photovoltaic power plants. Long-term forecasts predict photovoltaic power output for several years, and the results are used not only for solar resource assessment and the planning and construction of photovoltaic power plants but also for formulating long-term power generation plans.

[0005] Existing photovoltaic power prediction methods are mainly single-value predictions. Although single-value predictions can provide the expected value of future photovoltaic power, they are difficult to fully describe the fluctuation of photovoltaic power.

[0006] Probabilistic forecasting can provide future photovoltaic power values ​​and their probability distribution, offering more comprehensive predictive information and overcoming the limitations of single-value forecasting. Especially under complex weather conditions, probabilistic photovoltaic power forecasting results are more advantageous than single-value forecasting results. Summary of the Invention

[0007] The present invention aims to at least partially solve one of the technical problems in the related art.

[0008] Therefore, the first objective of this invention is to propose an ultra-short-term photovoltaic power range prediction method, which solves the problem of the effectiveness of photovoltaic power prediction, improves the security of photovoltaic integration into the power system, and at the same time, high-precision photovoltaic power prediction can increase the competitive advantage of photovoltaic power plants in the power market decision-making.

[0009] The second objective of this invention is to provide an ultra-short-term photovoltaic power range prediction device.

[0010] The third objective of this invention is to propose an ultra-short-term photovoltaic power range prediction system.

[0011] The fourth objective of this invention is to provide a computer device.

[0012] To achieve the above objectives, the first aspect of this invention proposes a method for predicting ultra-short-term photovoltaic power ranges, comprising:

[0013] S101, acquire photovoltaic power plant operation data and meteorological data; S102, input the photovoltaic power plant operation data and meteorological data into the trained spatiotemporal feedforward neural network model to predict and obtain the first single-value result; S103, input the first single-value result into the probability distribution function, and calculate the second single-value result based on the probability distribution function; S104, solve the first single-value result and the second single-value result into a system of equations, and calculate and output the photovoltaic power range prediction result based on the system of equations.

[0014] In addition, the ultra-short-term photovoltaic power range prediction method according to the above embodiments of the present invention may also have the following additional technical features:

[0015] Furthermore, in one embodiment of the present invention, the method further includes: acquiring sample data of photovoltaic power station operation data and meteorological data; performing data preprocessing and data cleaning to remove abnormal data from the sample data to obtain preprocessed data; wherein, the data cleaning includes removing shutdown data, sensor abnormal data, and power abnormal data; and selecting a preset number of feature data based on the preprocessed data to train a spatiotemporal feedforward neural network model to obtain the trained spatiotemporal feedforward neural network model.

[0016] Furthermore, in one embodiment of the present invention, the acquisition of photovoltaic power station operation data and meteorological data includes: acquiring power, shortwave radiation, longwave radiation, wind speed, wind direction, cloud cover, air pressure, precipitation, atmospheric temperature and relative humidity.

[0017] Furthermore, in one embodiment of the present invention, the spatiotemporal feedforward neural network model with the ANN structure is trained by calculating the i-th neuron in the hidden layer using a sinusoidal activation function g, expressed as:

[0018]

[0019]

[0020] Where j is the input parameter, w p,i b is the weight of the i-th neuron in each dimension. i It is the deviation threshold of the i-th neuron.

[0021] Furthermore, in one embodiment of the present invention, the i-th neuron of the output layer is calculated based on the sinusoidal activation function using a linear activation function to obtain the first single-valued result:

[0022]

[0023] Furthermore, in one embodiment of the present invention, the second single-valued result predicted based on the probability distribution function is expressed as:

[0024] y = h(x; γ) * )+e (3)

[0025] Where e is the difference between the predicted value and the actual value, γ * It is a set of parameters in the real number set that best reflects solar radiation;

[0026] The expression for the first single-valued result obtained by prediction using the trained spatiotemporal feedforward neural network model is as follows:

[0027]

[0028] in, It is γ * The result approximated by the LM algorithm is h, which is equation (2);

[0029] Using equations (3), (4) and e, the expression obtained is:

[0030]

[0031] in,

[0032] The transformed equation (5) is expressed as follows:

[0033]

[0034]

[0035]

[0036] in, It is a Jacobian matrix, R is the number of columns in matrix J, and γ is the number of columns in matrix J. * The number of parameters in the matrix, K, is the number of rows in the J matrix, and is used to calculate... The number of database samples.

[0037] Furthermore, in one embodiment of the present invention, the equations in the expression of equation (5) are combined to calculate the interval prediction result:

[0038]

[0039] Among them, u 2 It is σ 2 Unbiased estimate:

[0040]

[0041] Define random parameters:

[0042]

[0043] At a confidence level of 100(1-α)%, The interval prediction is:

[0044]

[0045] Where α∈[0,1] is the user's error rate.

[0046] The ultra-short-term photovoltaic power range prediction method of this invention is used to solve the problem of the effectiveness of photovoltaic power prediction, improve the security of photovoltaic integration into the power system, and at the same time, high-precision photovoltaic power prediction can increase the competitive advantage of photovoltaic power plants in the power market decision-making.

[0047] In a second aspect, the present invention provides an ultra-short-term photovoltaic power range prediction device, comprising:

[0048] The system comprises the following modules: a data acquisition module for acquiring photovoltaic power plant operation data and meteorological data; a first calculation module for inputting the photovoltaic power plant operation data and meteorological data into a trained spatiotemporal feedforward neural network model to predict and obtain a first single-value result; a second calculation module for inputting the first single-value result into a probability distribution function and calculating a second single-value result based on the probability distribution function; and a result prediction module for simultaneously solving a system of equations with the first and second single-value results and calculating and outputting a photovoltaic power range prediction result based on the system of equations.

[0049] The ultra-short-term photovoltaic power range prediction device of this invention solves the problem of the effectiveness of photovoltaic power prediction, improves the security of photovoltaic integration into the power system, and at the same time, the high-precision photovoltaic power prediction can increase the competitive advantage of photovoltaic power plants in the power market decision-making.

[0050] In a third aspect, the present invention proposes an ultra-short-term photovoltaic power range prediction system, comprising:

[0051] The system consists of a data acquisition module, a gateway module, a local database, and a prediction calculation module, which are connected in sequence. The prediction calculation module is used to: retrieve photovoltaic power plant operation data and meteorological data from the database; input the photovoltaic power plant operation data and meteorological data into a trained spatiotemporal feedforward neural network model to predict and obtain a first single-value result; input the first single-value result into a probability distribution function, and calculate a second single-value result based on the probability distribution function; and solve a system of equations combining the first single-value result and the second single-value result, and calculate and output the photovoltaic power range prediction result based on the system of equations.

[0052] In a fourth aspect, the present invention provides a computer device comprising: a processor and a memory;

[0053] The processor reads executable program code stored in memory to run a program corresponding to the executable program code, so as to implement the ultra-short-term photovoltaic power range prediction method as described in the first aspect embodiment.

[0054] Additional aspects and advantages of the invention will be set forth in part in the description which follows, and in part will be obvious from the description, or may be learned by practice of the invention. Attached Figure Description

[0055] The above and / or additional aspects and advantages of the present invention will become apparent and readily understood from the following description of the embodiments taken in conjunction with the accompanying drawings, wherein:

[0056] Figure 1 This is a flowchart of an ultra-short-term photovoltaic power range prediction method according to an embodiment of the present invention;

[0057] Figure 2 This is a schematic diagram of the structure of an ultra-short-term photovoltaic power range prediction device according to an embodiment of the present invention;

[0058] Figure 3 This is a schematic diagram of the ultra-short-term photovoltaic power range prediction system according to an embodiment of the present invention;

[0059] Figure 4 This is a schematic diagram of a computer device according to an embodiment of the present invention. Detailed Implementation

[0060] It should be noted that, unless otherwise specified, the embodiments and features described in this application can be combined with each other. The present invention will now be described in detail with reference to the accompanying drawings and embodiments.

[0061] To enable those skilled in the art to better understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present invention.

[0062] The following description, with reference to the accompanying drawings, outlines a method, apparatus, system, and equipment for predicting ultra-short-term photovoltaic power ranges according to embodiments of the present invention.

[0063] Figure 1 This is a flowchart of an ultra-short-term photovoltaic power range prediction method according to an embodiment of the present invention.

[0064] like Figure 1 As shown, the method includes, but is not limited to, the following steps:

[0065] S101, acquires photovoltaic power station operation data and meteorological data.

[0066] Specifically, the system acquires operational and meteorological data from photovoltaic power plants, such as power output, shortwave radiation, longwave radiation, wind speed, wind direction, cloud cover, air pressure, precipitation, atmospheric temperature, and relative humidity. Then, it performs data cleaning and preprocessing to remove data related to shutdowns, abnormal sensor data, and abnormal power output.

[0067] S102, input the photovoltaic power station operation data and meteorological data into the trained spatiotemporal feedforward neural network model to predict and obtain the first single-value result.

[0068] It is understandable that key features are selected for model training to construct a spatiotemporal feedforward neural network model.

[0069] As an example, the spatiotemporal feedforward neural network model is an ANN structure that relies on a combination of several parameters, which are selected through sensitivity analysis. During model training, the output is calculated at the i-th neuron in the hidden layer using a sinusoidal activation function g:

[0070]

[0071]

[0072] Where j is the input parameter, w p,ib is the weight of the i-th neuron in each dimension. i This is the bias threshold for the i-th neuron. Similarly, the i-th neuron in the output layer, i.e., the predicted single value, is calculated using a linear activation function:

[0073]

[0074] The trained spatiotemporal feedforward neural network model is read to predict the acquired meteorological data and obtain the first predicted value.

[0075] S103, input the first single-value result into the probability distribution function, and calculate the second single-value result based on the probability distribution function.

[0076] Specifically, based on the probability distribution function, the predicted second single-value result can be expressed as:

[0077] y = h(x; γ) * )+e (3)

[0078] Where e is the difference between the predicted value and the actual value, γ * It is a set of parameters in the real number set that best reflects solar radiation, and can be approximated by the LM algorithm.

[0079] Therefore, the first single-value result obtained through model prediction can be expressed as:

[0080]

[0081] in, It is γ * The result obtained by approximating the h equation using the LM algorithm is the previously mentioned equation (2).

[0082] S104: Combine the first single-valued result and the second single-valued result into a system of equations, and calculate and output the photovoltaic power range prediction result based on the system of equations.

[0083] Specifically, through equations (3) and (4), e can be expressed as:

[0084]

[0085] in,

[0086] Equation (5) can be further expressed as:

[0087]

[0088]

[0089]

[0090] in, It is a Jacobian matrix, R is the number of columns in matrix J, and γ is the number of columns in matrix J. * The number of parameters in the matrix; K is the number of rows in the J matrix, used to calculate... The number of database samples. Combining the two equations above, we can obtain the following result:

[0091]

[0092] u 2 It is σ 2 The unbiased estimate,

[0093] Finally, define random parameters. The student follows a t-distribution with KR degrees of freedom.

[0094] Therefore, it can be concluded that at a confidence level of 100(1-α)%... The interval prediction is:

[0095]

[0096] Where α∈[0,1] is the user's error rate.

[0097] Finally, output the interval prediction results.

[0098] The ultra-short-term photovoltaic power range prediction method according to embodiments of the present invention solves the problem of the effectiveness of photovoltaic power prediction, improves the security of photovoltaic integration into the power system, and at the same time, high-precision photovoltaic power prediction can increase the competitive advantage of photovoltaic power plants in the power market decision-making.

[0099] To achieve the above embodiments, such as Figure 2 As shown, this embodiment also provides an ultra-short-term photovoltaic power range prediction device 10, which includes: a data acquisition module 100, a first calculation module 200 and a second calculation module 300, and a result prediction module 400.

[0100] Data acquisition module 100 is used to acquire photovoltaic power station operation data and meteorological data;

[0101] The first calculation module 200 is used to input photovoltaic power station operation data and meteorological data into a trained spatiotemporal feedforward neural network model to predict and obtain the first single-value result.

[0102] The second calculation module 300 is used to input the first single-value result into the probability distribution function, and calculate the second single-value result based on the probability distribution function.

[0103] The result prediction module 400 is used to combine the first single-value result and the second single-value result into a system of equations, and calculate and output the photovoltaic power range prediction result based on the system of equations.

[0104] The ultra-short-term photovoltaic power range prediction device according to the present invention solves the problem of the effectiveness of photovoltaic power prediction, improves the security of photovoltaic integration into the power system, and at the same time, the high-precision photovoltaic power prediction can increase the competitive advantage of photovoltaic power plants in the power market decision-making.

[0105] To achieve the above embodiments, such as Figure 3 As shown, this embodiment also provides an ultra-short-term photovoltaic power range prediction system 20, including:

[0106] The data acquisition module 201, gateway module 202, local database 203, and prediction calculation module 204 are connected in sequence.

[0107] Prediction calculation module 204 is used for:

[0108] Retrieve photovoltaic power station operation data and meteorological data from local database 203;

[0109] The photovoltaic power station operation data and meteorological data are input into the trained spatiotemporal feedforward neural network model to obtain the first single value result;

[0110] The first single-value result is input into the probability distribution function, and the second single-value result is obtained by calculating the first single-value result based on the probability distribution function;

[0111] The first and second single-value results are combined into a system of equations, and the photovoltaic power range prediction results are calculated and output based on the system of equations.

[0112] The ultra-short-term photovoltaic power range prediction system according to embodiments of the present invention solves the problem of the effectiveness of photovoltaic power prediction, improves the security of photovoltaic integration into the power system, and at the same time, high-precision photovoltaic power prediction can increase the competitive advantage of photovoltaic power plants in the power market decision-making.

[0113] To implement the methods of the above embodiments, the present invention also provides a computer device, such as... Figure 4 As shown, the computer device 600 includes a memory 601 and a processor 602; wherein, the processor 602 reads executable program code stored in the memory 601 to run a program corresponding to the executable program code, so as to implement the various steps of the method described above.

[0114] In the description of this specification, the references to terms such as "one embodiment," "some embodiments," "example," "specific example," or "some examples," etc., refer to specific features, structures, materials, or characteristics described in connection with that embodiment or example, which are included in at least one embodiment or example of this application. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples. Moreover, without contradiction, those skilled in the art can combine and integrate the different embodiments or examples described in this specification, as well as the features of different embodiments or examples.

[0115] Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include at least one of that feature. In the description of this application, "multiple" means at least two, such as two, three, etc., unless otherwise explicitly specified.

[0116] Any process or method description in the flowchart or otherwise herein can be understood as representing a module, segment, or portion of code comprising one or more executable instructions for implementing custom logic functions or processes, and the scope of the preferred embodiments of this application includes additional implementations in which functions may be performed not in the order shown or discussed, including substantially simultaneously or in reverse order depending on the functions involved, as should be understood by those skilled in the art to which embodiments of this application pertain.

[0117] The logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as a ordered list of executable instructions for implementing logical functions, and can be embodied in any computer-readable medium for use by, or in conjunction with, an instruction execution system, apparatus, or device (such as a computer-based system, a processor-included system, or other system that can fetch and execute instructions from, an instruction execution system, apparatus, or device). For the purposes of this specification, "computer-readable medium" can be any means that can contain, store, communicate, propagate, or transmit programs for use by, or in conjunction with, an instruction execution system, apparatus, or device. More specific examples (a non-exhaustive list) of computer-readable media include: an electrical connection having one or more wires (electronic device), a portable computer disk drive (magnetic device), random access memory (RAM), read-only memory (ROM), erasable and editable read-only memory (EPROM or flash memory), fiber optic devices, and portable optical disc read-only memory (CDROM). Alternatively, the computer-readable medium may be paper or other suitable media on which the program can be printed, since the program can be obtained electronically, for example, by optically scanning the paper or other medium, followed by editing, interpreting, or otherwise processing as necessary, and then stored in a computer memory.

[0118] It should be understood that various parts of this application can be implemented using hardware, software, firmware, or a combination thereof. In the above embodiments, multiple steps or methods can be implemented using software or firmware stored in memory and executed by a suitable instruction execution system. For example, if implemented in hardware as in another embodiment, it can be implemented using any one or a combination of the following techniques known in the art: discrete logic circuits having logic gates for implementing logical functions on data signals, application-specific integrated circuits (ASICs) having suitable combinational logic gates, programmable gate arrays (PGAs), field-programmable gate arrays (FPGAs), etc.

[0119] Those skilled in the art will understand that all or part of the steps of the methods in the above embodiments can be implemented by a program instructing related hardware. The program can be stored in a computer-readable storage medium, and when executed, the program includes one or a combination of the steps of the method embodiments.

[0120] Furthermore, the functional units in the various embodiments of this application can be integrated into a processing module, or each unit can exist physically separately, or two or more units can be integrated into a module. The integrated module can be implemented in hardware or as a software functional module. If the integrated module is implemented as a software functional module and sold or used as an independent product, it can also be stored in a computer-readable storage medium.

[0121] The storage medium mentioned above can be a read-only memory, a disk, or an optical disk, etc. Although embodiments of this application have been shown and described above, it is understood that the above embodiments are exemplary and should not be construed as limiting this application. Those skilled in the art can make changes, modifications, substitutions, and variations to the above embodiments within the scope of this application.

Claims

1. A method for predicting ultra-short-term photovoltaic power range, characterized in that, Includes the following steps: S101, acquires photovoltaic power station operation data and meteorological data; S102, input the photovoltaic power station operation data and meteorological data into the trained spatiotemporal feedforward neural network model to predict and obtain the first single-value result; S103, input the first single-value result into the probability distribution function, and calculate the second single-value result based on the probability distribution function; S104, combine the first single-value result and the second single-value result into a system of equations, and calculate and output the photovoltaic power range prediction result based on the system of equations; The method further includes: Obtain sample data of photovoltaic power station operation data and meteorological data; The sample data is preprocessed and cleaned to remove abnormal data, resulting in preprocessed data; wherein, the data cleaning includes removing shutdown data, abnormal sensor data, and abnormal power data. Based on the preprocessed data, a preset amount of feature data is selected to train a spatiotemporal feedforward neural network model to obtain the trained spatiotemporal feedforward neural network model. The spatiotemporal feedforward neural network model with the ANN structure is trained by calculating the i-th neuron in the hidden layer using a sinusoidal activation function g, expressed as: （1） Where j is the input parameter, It represents the weight of the i-th neuron in each dimension. It is the deviation threshold of the i-th neuron; The i-th neuron of the output layer is calculated using the sinusoidal activation function in the form of a linear activation function to obtain the first single-value result: （2）； The second single-valued result predicted based on the probability distribution function is expressed as follows: （3） Where e is the difference between the predicted value and the actual value. It is a set of parameters in the real number set that best reflects solar radiation; The expression for the first single-valued result obtained by prediction using the trained spatiotemporal feedforward neural network model is as follows: （4） in, yes The result approximated by the LM algorithm is h, which is equation (2). Using equations (3), (4) and e, the expression obtained is: （5） in, The transformed equation (5) is expressed as follows: in, It is a Jacobian matrix, and R is the number of columns in matrix J. The number of parameters in the matrix, K, is the number of rows in the J matrix, and is used to calculate... The number of database samples; Combining the equations in expression (5), the interval prediction result is calculated: in, yes Unbiased estimate: Define random parameters: Get confidence level Down The interval prediction is: in, It is the error rate of user selection. .

2. The method according to claim 1, characterized in that, The acquisition of photovoltaic power station operation data and meteorological data includes: Acquire power, shortwave radiation, longwave radiation, wind speed, wind direction, cloud cover, air pressure, precipitation, atmospheric temperature, and relative humidity.

3. A device for predicting ultra-short-term photovoltaic power range, characterized in that, The device is used to implement the ultra-short-term photovoltaic power range prediction method as described in claim 1, and the device includes: The data acquisition module is used to acquire photovoltaic power station operation data and meteorological data; The first calculation module is used to input the photovoltaic power station operation data and meteorological data into a trained spatiotemporal feedforward neural network model to predict and obtain the first single-value result. The second calculation module is used to input the first single-value result into a probability distribution function, and calculate the second single-value result based on the probability distribution function. The result prediction module is used to combine the first single-value result and the second single-value result into a system of equations, and calculate and output the photovoltaic power range prediction result based on the system of equations.

4. A short-term photovoltaic power range prediction system, the system being used to implement the short-term photovoltaic power range prediction method as described in claim 1, characterized in that, include: The data acquisition module, gateway module, local database, and prediction calculation module are connected in sequence. The prediction calculation module is used for: Retrieve photovoltaic power station operation data and meteorological data from the database; The photovoltaic power station operation data and meteorological data are input into a trained spatiotemporal feedforward neural network model to predict and obtain the first single-value result. The first single-value result is input into the probability distribution function, and the second single-value result is calculated based on the probability distribution function. The first single-value result and the second single-value result are combined into a system of equations, and the photovoltaic power range prediction result is calculated and output based on the system of equations.

5. A computer device, characterized in that, Including processor and memory; The processor reads executable program code stored in the memory to run a program corresponding to the executable program code, so as to implement the ultra-short-term photovoltaic power range prediction method as described in any one of claims 1-2.

Images