Medium and long term wind and light resource prediction method and system considering climate remote correlation factors

Abstract

The invention belongs to the technical field of electric power meteorology, and provides a medium and long term wind and light resource prediction method and system considering climate remote correlation factors, and the method comprises the steps: obtaining global climate index historical data, and carrying out the standardization processing of the data, and obtaining a standardized climate index; for the historical resource sequence of the target station, calculating the mutual information value of the standardized climate index, and constructing a forecasting factor set by taking the lag time corresponding to the maximum value of the mutual information as the optimal early warning window period; extracting wind and light resource measured data of the same period in historical years to construct a reference probability density function; and taking the reference probability density function as prior distribution, combining a preset dynamic mode prediction result as a likelihood function, and outputting a prediction result containing a deterministic value and a confidence interval. By means of the characteristic that ocean signals change slowly, the method breaks through the 15-day prediction limit of an atmospheric mode, effective trend prediction from the quarterly level to the annual level is successfully achieved, and the requirement for medium and long term resource evaluation is met.

Description

Technical Field

[0001] This invention belongs to the field of power meteorology technology, specifically relating to a method and system for predicting medium- and long-term wind and solar resources that takes into account climate teleconnection factors. Background Technology

[0002] As renewable energy penetration in the power system increases, medium- and long-term wind and solar resource assessments are crucial for power trading, maintenance planning, and asset valuation.

[0003] Existing medium- and long-term wind and solar resource forecasting technologies have significant limitations, resulting in large interannual errors in resource forecasting in regions with significant monsoon climates.

[0004] The current mainstream solutions fall into two categories: one is numerical weather prediction (NWP) based on global climate models (GCM), and the other is traditional statistical methods based on autocorrelation of local historical data, such as ARIMA and Markov chains. In addition, there are a few qualitative applied studies on climate factors such as El Niño (ENSO) and North Atlantic Oscillation (NAO) in academia.

[0005] Numerical weather forecasting suffers from the decay problem of initial field error amplifying over time through integration. After 15 days, the daily deterministic forecasting capability decreases significantly, making it unable to directly guide power generation forecasting at power plants. Traditional statistical methods suffer from memorylessness, relying solely on local data and failing to capture interannual fluctuations and abrupt changes in wind speed and irradiance controlled by teleconnections of air-sea interactions. Climate factors are applied crudely, lacking quantitative engineering models that consider lag time and nonlinear response intensity for specific power plants. Overall, there is a lack of a systematic method to effectively downscale and map large-scale, long-period climate signals to micro-level power plant resource fluctuations, resulting in extremely large interannual errors in resource forecasting in regions with significant monsoon climates. Summary of the Invention

[0006] The purpose of this invention is to provide a medium- and long-term wind and solar resource forecasting method and system that takes into account climate teleconnection factors, so as to solve the problems of poor forecasting ability and large forecasting error in existing technologies.

[0007] To achieve the above objectives, the present invention adopts the following technical solution: In a first aspect, the present invention provides a method for predicting medium- and long-term wind and solar resources that takes into account climate teleconnection factors, including: Obtain historical data of the global climate index, and then standardize the data to obtain the standardized climate index; For the historical resource sequence of the target site, the mutual information value of the standardized climate index is calculated, and the lag time corresponding to the maximum mutual information value is used as the optimal early warning window period to construct a forecast factor set; The current climate state vector is obtained based on the forecast factor set. The historical year with the closest Euclidean distance to the current state is searched in the historical database. The measured wind and solar resources data of the same period in the historical year are extracted to construct the benchmark probability density function. Using the baseline probability density function as the prior distribution and combining it with the preset dynamic model prediction results as the likelihood function, the prediction results containing deterministic values ​​and confidence intervals are output.

[0008] Furthermore, the step of obtaining historical global climate index data and standardizing the data to obtain a standardized climate index includes: Get the Global Climate Index Historical data, constructing a feature pool ; For feature pool The data is detrended and standardized.

[0009] in, The standardized climate index. and These represent the mean and standard deviation under a sliding window.

[0010] Furthermore, feature pool The characteristics include El Niño-Southern Oscillation (ENSO), North Atlantic Oscillation (NAO), Pacific Decadal Oscillation (PDO), Indian Ocean Dipole Oscillation (IOD), and Arctic Oscillation (AO).

[0011] Furthermore, for the historical resource sequence of the target site, the mutual information value of the standardized climate index is calculated, and the lag time corresponding to the maximum mutual information value is used as the optimal early warning window period to construct a forecast factor set, including: The maximum correlation time delay search algorithm based on mutual information is used to target the historical resource sequence of the site. Calculate its relationship with various climate factors Different lag times , Mutual information value under the moon :

[0012] Select The largest value As the optimal early warning window for this factor at this station, a forecast factor set is constructed, in which... for, for, for.

[0013] Furthermore, the step of obtaining the current climate state vector based on the forecast factor set, searching the historical database for the historical year with the closest Euclidean distance to the current state, and extracting measured wind and solar resource data from the same period of the historical year to construct a baseline probability density function includes: Key factors selected based on the forecast factor set and its rate of change Constitutes the current climate state vector :

[0014] in For the current moment, This represents the total number of key climate factors selected.

[0015] Search the historical database for the Euclidean algorithm that is closest to the current state. Historical years:

[0016] Extract measured wind / solar resource data from the same period in similar years to construct a baseline probability density function (PDF) for prediction, where: For the current year (prediction starting point) of the [number]th State values ​​of each climate factor For the first The first historical year in the same period State values ​​of each climate factor For the current year's climate state and the first The Euclidean distance between historical years indicates that the climate modes of the two years are more similar. The total number of key climate factor characteristics involved in the calculation. Index of climate factor characteristics (from 1 to ), For the first The weighting coefficients of each climate factor, reflecting the importance of that factor to local resources, are used to calculate the weighted similarity. For the current year (prediction starting point) of the [number]th State values ​​of each climate factor For the first The first historical year in the same period The state values ​​of each climate factor.

[0017] Furthermore, the step of using the baseline probability density function as a prior distribution and combining it with the preset dynamic mode prediction results as a likelihood function to output a prediction result containing deterministic values ​​and confidence intervals includes: A prediction equation is established, in which a nonlinear response term of climate factors is introduced. :

[0018] in To predict resource quantity, This is the climatological mean. This is a non-linear mapping function fitted using machine learning. , These represent the proportions or contributions of climatological mean, nonlinear terms of climate factors, and dynamic model forecast terms in the final prediction results. This represents specific climate factor values ​​(such as the ENSO index), which here indicate the key climate predictors input into the model. This represents the seasonal dynamic model forecast (Seasonal NWP). It is a direct numerical output from global climate models (such as CFS), serving as the physical background field input. This refers to the residual term or the random error term.

[0019] The output includes deterministic values ​​and confidence intervals based on the distribution of similar historical years.

[0020] Secondly, the present invention provides a medium- to long-term wind and solar resource prediction system that considers climate teleconnection factors, comprising: The data acquisition module is used to acquire historical data of the global climate index and then standardize the data to obtain the standardized climate index. The calculation module is used to calculate the mutual information value of the standardized climate index for the historical resource sequence of the target site, and to construct the forecast factor set by taking the lag time corresponding to the maximum mutual information value as the optimal early warning window period. The function construction module is used to obtain the current climate state vector based on the forecast factor set, search the historical database for the historical year with the closest Euclidean distance to the current state, and extract the measured wind and solar resources data of the same period of the historical year to construct the benchmark probability density function. The prediction output module is used to take the baseline probability density function as the prior distribution and combine it with the preset dynamic mode prediction results as the likelihood function to output the prediction results containing deterministic values ​​and confidence intervals.

[0021] Furthermore, the step of obtaining historical global climate index data and standardizing the data to obtain a standardized climate index includes: Get the Global Climate Index Historical data, constructing a feature pool ; For feature pool The data is detrended and standardized.

[0022] in, The standardized climate index. and These represent the mean and standard deviation under a sliding window.

[0023] Furthermore, feature pool The characteristics include El Niño-Southern Oscillation (ENSO), North Atlantic Oscillation (NAO), Pacific Decadal Oscillation (PDO), Indian Ocean Dipole Oscillation (IOD), and Arctic Oscillation (AO).

[0024] Furthermore, for the historical resource sequence of the target site, the mutual information value of the standardized climate index is calculated, and the lag time corresponding to the maximum mutual information value is used as the optimal early warning window period to construct a forecast factor set, including: The maximum correlation time delay search algorithm based on mutual information is used to target the historical resource sequence of the site. Calculate its relationship with various climate factors Different lag times , Mutual information value under the moon :

[0025] Select The largest value As the optimal early warning window for this factor at this station, a forecast factor set is constructed, in which... Let be a joint probability distribution, representing the value of the local resource. And the climate factor values ​​are The probability of them happening simultaneously This represents the marginal probability distribution of local resource data, that is, considering the probability of resource data occurrence in isolation. This represents the marginal probability distribution of climate factor data, i.e., the probability of a climate factor value occurring when considered individually.

[0026] Furthermore, the step of obtaining the current climate state vector based on the forecast factor set, searching the historical database for the historical year with the closest Euclidean distance to the current state, and extracting measured wind and solar resource data from the same period of the historical year to construct a baseline probability density function includes: Key factors selected based on the forecast factor set and its rate of change Constitutes the current climate state vector :

[0027] in For the current moment, This represents the total number of key climate factors selected.

[0028] Search the historical database for the Euclidean algorithm that is closest to the current state. Historical years:

[0029] Extract measured wind / solar resource data from the same period in similar years to construct a baseline probability density function (PDF) for prediction, where: For the current year (prediction starting point) of the [number]th State values ​​of each climate factor For the first The first historical year in the same period State values ​​of each climate factor For the current year's climate state and the first The Euclidean distance between historical years indicates that the climate modes of the two years are more similar. The total number of key climate factor characteristics involved in the calculation. Index of climate factor characteristics (from 1 to ), For the first The weighting coefficients of each climate factor, reflecting the importance of that factor to local resources, are used to calculate the weighted similarity. For the current year (prediction starting point) of the [number]th State values ​​of each climate factor For the first The first historical year in the same period The state values ​​of each climate factor.

[0030] Furthermore, the step of using the baseline probability density function as a prior distribution and combining it with the preset dynamic mode prediction results as a likelihood function to output a prediction result containing deterministic values ​​and confidence intervals includes: A prediction equation is established, in which a nonlinear response term of climate factors is introduced. :

[0031] in To predict resource quantity, This is the climatological mean. This is a non-linear mapping function fitted using machine learning. , These represent the proportions or contributions of climatological mean, nonlinear terms of climate factors, and dynamic model forecast terms in the final prediction results. This represents specific climate factor values ​​(such as the ENSO index), which here indicate the key climate predictors input into the model. This represents the seasonal dynamic model forecast (Seasonal NWP). It is a direct numerical output from global climate models (such as CFS), serving as the physical background field input. This refers to the residual term or the random error term.

[0032] The output includes deterministic values ​​and confidence intervals based on the distribution of similar historical years.

[0033] Thirdly, the present invention provides a computer device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the steps of the medium- and long-term wind and solar resource prediction method considering climate teleconnection factors.

[0034] Fourthly, the present invention provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the steps of the medium- and long-term wind and solar resource prediction method considering climate teleconnection factors.

[0035] Compared with the prior art, the present invention has the following technical effects: This invention leverages the slow-changing nature of ocean signals to overcome the 15-day forecasting limit of atmospheric models, successfully achieving effective trend forecasting at the quarterly to annual levels and meeting the needs of medium- and long-term resource assessment.

[0036] This invention has strong physical interpretability, unlike black-box artificial intelligence (models), which can clearly provide a complete logical chain. This clear causal relationship is more easily accepted by power dispatchers.

[0037] This invention can effectively capture extreme scenarios. Through similar year analysis, it can accurately identify "extremely dry" or "extremely abundant" resource events under similar climatic backgrounds in history, providing key decision-making basis for energy supply planning and power trading risk hedging. Attached Figure Description

[0038] Figure 1 This is a flowchart of the present invention.

[0039] Figure 2 This is a logic block diagram of the present invention. Detailed Implementation

[0040] The present invention will be further described below with reference to the accompanying drawings: Example 1, please refer to Figure 1 This invention provides a method for predicting medium- and long-term wind and solar resources that considers climate teleconnection factors, including: Obtain historical data of the global climate index, and then standardize the data to obtain the standardized climate index; For the historical resource sequence of the target site, the mutual information value of the standardized climate index is calculated, and the lag time corresponding to the maximum mutual information value is used as the optimal early warning window period to construct a forecast factor set; The current climate state vector is obtained based on the forecast factor set. The historical year with the closest Euclidean distance to the current state is searched in the historical database. The measured wind and solar resources data of the same period in the historical year are extracted to construct the benchmark probability density function. Using the baseline probability density function as the prior distribution and combining it with the preset dynamic model prediction results as the likelihood function, the prediction results containing deterministic values ​​and confidence intervals are output.

[0041] This invention leverages the slowly varying characteristics of ocean signals to overcome the 15-day prediction limit of atmospheric models, achieving effective trend forecasting at the quarterly to annual levels. Through similar year analysis, it can effectively identify extreme drought or abundance events under similar historical climate conditions, providing crucial intelligence for energy supply security and power trading hedging.

[0042] Example 2, please refer to Figure 2 This invention provides a method for predicting medium- and long-term wind and solar resources that considers climate teleconnection factors, including: (1) Construct a multidimensional climate teleconnection factor feature library: Historical global climate index data were obtained from authoritative meteorological agencies (such as NOAA and ECMWF) to construct a feature pool. The characteristics include not only ENSO (Nino 3.4 index) and NAO (North Atlantic Oscillation), but also PDO (Pacific Decadal Oscillation), IOD (Indian Ocean Dipole), and AO (Arctic Oscillation). Detrending and standardizing the data helps to eliminate the masking effect of global warming trends on fluctuation signals.

[0043] in, The standardized climate index. and These represent the mean and standard deviation under a sliding window.

[0044] (2) Identification of dynamic time lags in specific regions: The transmission time of climatic factors varies in different regions (for example, the impact of ENSO on wind speed in eastern China may lag by 3-6 months). This invention proposes a maximum correlation time lag search algorithm based on mutual information. Historical resource sequence of the target site Calculate its relationship with various climate factors Different lag times ( Mutual information value under (month) :

[0045] Select The largest value This serves as the optimal early warning window for the station, thus constructing a set of forecast factors with physical significance. Among them... Let be a joint probability distribution, representing the value of the local resource. And the climate factor values ​​are The probability of them happening simultaneously This represents the marginal probability distribution of local resource data, that is, considering the probability of resource data occurrence in isolation. This represents the marginal probability distribution of climate factor data, i.e., the probability of a climate factor value occurring when considered individually.

[0046] (3) Based on similar year weighted prediction of climate mode classification, the traditional linear regression is abandoned and the "climate mode matching" approach is adopted: Key factors selected based on the forecast factor set and its rate of change Constitutes the current climate state vector :

[0047] in For the current moment, This represents the total number of key climate factors selected.

[0048] Search the historical database for the Euclidean algorithm that is closest to the current state. Historical years:

[0049] Extract measured wind / solar resource data from the same period in similar years to construct a baseline probability density function (PDF) for prediction, where: For the current year (prediction starting point) of the [number]th State values ​​of each climate factor For the first The first historical year in the same period State values ​​of each climate factor For the current year's climate state and the first The Euclidean distance between historical years indicates that the climate modes of the two years are more similar. The total number of key climate factor characteristics involved in the calculation. Index of climate factor characteristics (from 1 to ), For the first The weighting coefficients of each climate factor, reflecting the importance of that factor to local resources, are used to calculate the weighted similarity. For the current year (prediction starting point) of the [number]th State values ​​of each climate factor For the first The first historical year in the same period The state values ​​of each climate factor.

[0050] (4) Construct a Bayesian hierarchical probabilistic prediction model. The physical statistics obtained in step (3) are used as the prior distribution, and the prediction results of dynamic models (such as the US Climate Prediction System CFS) are used as the likelihood function to output the final resource anomaly prediction: A prediction equation is established, in which a nonlinear response term of climate factors is introduced. :

[0051] in To predict resource quantity, This is the climatological mean. For a non-linear mapping function fitted by machine learning (such as support vector regression, SVR); , These represent the proportions or contributions of climatological mean, nonlinear terms of climate factors, and dynamic model forecast terms in the final prediction results. This represents specific climate factor values ​​(such as the ENSO index), which here indicate the key climate predictors input into the model. This is the seasonal dynamic model forecast (Seasonal NWP). It is a direct numerical output from global climate models (such as CFS), serving as the physical background field input. This refers to the residual term or the random error term.

[0052] The output includes not only deterministic values, but also confidence intervals based on the distribution of similar historical years (such as probability values ​​for P75 and P90).

[0053] By calculating the mutual information or maximum correlation coefficient between the historical resource sequence of the target site and the global climate factor sequence within a sliding time window, the system automatically identifies the "optimal lag month" for each climate factor on that specific site and reconstructs the input feature set accordingly. This solves the problems of traditional methods that rigidly apply climate indices and ignore geographical location and transmission time differences, enabling precise factor selection based on a "one site, one policy" approach.

[0054] By utilizing the current multidimensional climate factor state vector (such as the combination of ENSO phase and NAO phase), the K nearest neighbor "similar climate years" are retrieved from the historical database. The measured distribution of resources in the same period of these similar years is used to construct the prediction probability density function (PDF) for future periods, instead of just outputting a single mean. Introducing the concept of "similar forecasting" from climatology into the field of new energy effectively quantifies the forecast uncertainty caused by the variability within the climate system.

[0055] Nonlinear regression models are constructed using climate teleconnection factors, specifically designed to predict and correct systematic biases in numerical climate models (such as the US climate prediction system CFSv2) in specific regions, particularly for second-order correction of precipitation / wind speed trend biases in monsoon marginal zones.

[0056] The core principle of this scheme is to leverage the "long memory" of the ocean-atmosphere system (such as the persistence of sea surface temperature anomalies) to overcome the problem that existing medium- and long-term wind and solar resource forecasts ignore large-scale climate forcing signals. By establishing a dynamic lag correlation between "climate factors and local resources," a closed-loop forecasting framework of "factor screening, time lag identification, mode matching, and probability correction" is constructed. This effectively "downscales" global-scale climate teleconnection signals (such as El Niño and the North Atlantic Oscillation) and maps them to the resource fluctuations of specific wind farms and photovoltaic power plants, thereby achieving accurate medium- and long-term forecasts and providing quantitative support for climate resilience planning of power systems.

[0057] Example 3, taking a certain offshore wind farm in the southeast coast of China as an example: (1) Data input: Input the monthly average wind speed of the wind farm's meteorological tower over the past 20 years, and the Nino3.4 index, AO index and other factors during the same period. (2) Time lag analysis: The system calculation found that the Nino3.4 index is significantly negatively correlated with the wind speed in the area, and the optimal lag time is 5 months (i.e., mainly affected by the sea surface temperature in the previous winter). (3) Prediction execution: Assuming that it is currently February, the Nino3.4 index is monitored to be continuously high (strong El Niño). The system automatically matches the year following strong El Niño such as 1998 and 2016 as similar years. (4) Result output: The system prompts "It is expected that the average wind speed of the wind farm from July to September will be 10%-15% lower than the normal level, and the probability of consecutive calm days will increase by 30%", and suggests that the operator adjust the power generation plan in advance and arrange unit maintenance.

[0058] In another embodiment of the present invention, a medium- to long-term wind and solar resource prediction system considering climate teleconnection factors is provided, which can be used to implement the above-mentioned medium- to long-term wind and solar resource prediction method considering climate teleconnection factors. Specifically, the system includes: The data acquisition module is used to acquire historical data of the global climate index and then standardize the data to obtain the standardized climate index. The calculation module is used to calculate the mutual information value of the standardized climate index for the historical resource sequence of the target site, and to construct the forecast factor set by taking the lag time corresponding to the maximum mutual information value as the optimal early warning window period. The function construction module is used to obtain the current climate state vector based on the forecast factor set, search the historical database for the historical year with the closest Euclidean distance to the current state, and extract the measured wind and solar resources data of the same period of the historical year to construct the benchmark probability density function. The prediction output module is used to take the baseline probability density function as the prior distribution and combine it with the preset dynamic mode prediction results as the likelihood function to output the prediction results containing deterministic values ​​and confidence intervals.

[0059] The module division in this embodiment of the invention is illustrative and represents only one logical functional division. In actual implementation, other division methods may be used. Furthermore, the functional modules in the various embodiments of the invention can be integrated into a single processor, exist as separate physical entities, or be integrated into a single module. The integrated modules described above can be implemented in hardware or as software functional modules.

[0060] In another embodiment of the present invention, a computer device is provided, comprising a processor and a memory. The memory stores a computer program, which includes program instructions. The processor executes the program instructions stored in the computer storage medium. The processor may be a Central Processing Unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. It is the computing and control core of the terminal, suitable for implementing one or more instructions, specifically suitable for loading and executing one or more instructions in the computer storage medium to achieve a corresponding method flow or corresponding function. The processor described in this embodiment of the present invention can be used for the operation of a medium- to long-term wind and solar resource prediction method considering climate teleconnection factors.

[0061] In another embodiment of the present invention, a storage medium is provided, specifically a computer-readable storage medium (Memory), which is a memory device in a computer device used to store programs and data. It is understood that the computer-readable storage medium here can include both the built-in storage medium in the computer device and extended storage media supported by the computer device. The computer-readable storage medium provides storage space that stores the terminal's operating system. Furthermore, the storage space also stores one or more instructions suitable for loading and execution by a processor. These instructions can be one or more computer programs (including program code). It should be noted that the computer-readable storage medium here can be high-speed RAM or non-volatile memory, such as at least one disk storage device. The processor can load and execute one or more instructions stored in the computer-readable storage medium to implement the corresponding steps of the medium- and long-term wind and solar resource prediction method considering climate teleconnection factors in the above embodiments.

[0062] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0063] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0064] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0065] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0066] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit it. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that modifications or equivalent substitutions can still be made to the specific implementation of the present invention. Any modifications or equivalent substitutions 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 medium- to long-term wind and solar resource prediction method considering climate teleconnection factors, characterized in that, include: Obtain historical data of the global climate index, and then standardize the data to obtain the standardized climate index; For the historical resource sequence of the target site, the mutual information value of the standardized climate index is calculated, and the lag time corresponding to the maximum mutual information value is used as the optimal early warning window period to construct a forecast factor set; The current climate state vector is obtained based on the forecast factor set. The historical year with the closest Euclidean distance to the current state is searched in the historical database. The measured wind and solar resources data of the same period in the historical year are extracted to construct the benchmark probability density function. Using the baseline probability density function as the prior distribution and combining it with the preset dynamic model prediction results as the likelihood function, the prediction results containing deterministic values ​​and confidence intervals are output.

2. The method for predicting medium- and long-term wind and solar resources considering climate teleconnection factors according to claim 1, characterized in that, The process of obtaining historical global climate index data and standardizing the data to obtain a standardized climate index includes: Get the Global Climate Index Historical data, constructing a feature pool ; For feature pool The data is detrended and standardized. in, The standardized climate index. and These represent the mean and standard deviation under a sliding window.

3. The method for predicting medium- and long-term wind and solar resources considering climate teleconnection factors according to claim 2, characterized in that, Feature pool The characteristics include El Niño-Southern Oscillation (ENSO), North Atlantic Oscillation (NAO), Pacific Decadal Oscillation (PDO), Indian Ocean Dipole Oscillation (IOD), and Arctic Oscillation (AO).

4. The method for predicting medium- and long-term wind and solar resources considering climate teleconnection factors according to claim 1, characterized in that, The process involves calculating the mutual information value of a standardized climate index for the historical resource sequence of the target site, using the lag time corresponding to the maximum mutual information value as the optimal early warning window, and constructing a forecast factor set, including: The maximum correlation time delay search algorithm based on mutual information is used to target the historical resource sequence of the site. Calculate its relationship with various climate factors Different lag times , Mutual information value under the moon : Select The largest value As the optimal early warning window for this factor at this station, a forecast factor set is constructed, in which... Let be a joint probability distribution, representing the value of the local resource. And the climate factor values ​​are The probability of them happening simultaneously This represents the marginal probability distribution of local resource data, that is, considering the probability of resource data occurrence in isolation. This represents the marginal probability distribution of climate factor data, i.e., the probability of a climate factor value occurring when considered individually.

5. The method for predicting medium- and long-term wind and solar resources considering climate teleconnection factors according to claim 1, characterized in that, The process of obtaining the current climate state vector based on the forecast factor set, searching the historical database for the historical year with the closest Euclidean distance to the current state, and extracting measured wind and solar resource data from the same period of the historical year to construct a baseline probability density function includes: Key factors selected based on the forecast factor set and its rate of change Constitutes the current climate state vector : in For the current moment, The total number of key climate factors selected; Search the historical database for the Euclidean algorithm that is closest to the current state. Historical years: Extract measured wind / solar resource data from the same period in similar years to construct a baseline probability density function (PDF) for prediction, where: For the current year (prediction starting point) of the [number]th State values ​​of each climate factor For the first The first historical year in the same period State values ​​of each climate factor For the current year's climate state and the first The Euclidean distance between historical years indicates that the climate modes of the two years are more similar. The total number of key climate factor characteristics involved in the calculation. An index for climate factor characteristics, from 1 to , For the first The weighting coefficients of each climate factor, reflecting the importance of that factor to local resources, are used to calculate the weighted similarity. For the current year, the State values ​​of each climate factor For the first The first historical year in the same period The state values ​​of each climate factor.

6. The method for predicting medium- and long-term wind and solar resources considering climate teleconnection factors according to claim 1, characterized in that, The process of using a baseline probability density function as a prior distribution and combining it with a preset dynamic mode prediction result as a likelihood function to output a prediction result containing deterministic values ​​and confidence intervals includes: A prediction equation is established, in which a nonlinear response term of climate factors is introduced. : in To predict resource quantity, This is the climatological mean. This is a non-linear mapping function fitted using machine learning. , These represent the proportions or contributions of climatological mean, nonlinear terms of climate factors, and dynamic model forecast terms in the final prediction results. This represents specific climate factor values ​​(such as the ENSO index), which here indicate the key climate predictors input into the model. These are values ​​predicted by seasonal dynamics models; It is a residual term or a random error term; The output includes deterministic values ​​and confidence intervals based on the distribution of similar historical years.

7. A medium- to long-term wind and solar resource forecasting system considering climate teleconnection factors, characterized in that, include: The data acquisition module is used to acquire historical data of the global climate index and then standardize the data to obtain the standardized climate index. The calculation module is used to calculate the mutual information value of the standardized climate index for the historical resource sequence of the target site, and to construct the forecast factor set by taking the lag time corresponding to the maximum mutual information value as the optimal early warning window period. The function construction module is used to obtain the current climate state vector based on the forecast factor set, search the historical database for the historical year with the closest Euclidean distance to the current state, and extract the measured wind and solar resources data of the same period of the historical year to construct the benchmark probability density function. The prediction output module is used to take the baseline probability density function as the prior distribution and combine it with the preset dynamic mode prediction results as the likelihood function to output the prediction results containing deterministic values ​​and confidence intervals.

8. The medium- and long-term wind and solar resource forecasting system considering climate teleconnection factors according to claim 7, characterized in that, The process of obtaining historical global climate index data and standardizing the data to obtain a standardized climate index includes: Get the Global Climate Index Historical data, constructing a feature pool ; For feature pool The data is detrended and standardized. in, The standardized climate index. and These represent the mean and standard deviation under a sliding window.

9. The medium- and long-term wind and solar resource forecasting system considering climate teleconnection factors according to claim 8, characterized in that, Feature pool The characteristics include El Niño-Southern Oscillation (ENSO), North Atlantic Oscillation (NAO), Pacific Decadal Oscillation (PDO), Indian Ocean Dipole Oscillation (IOD), and Arctic Oscillation (AO).

10. The medium- and long-term wind and solar resource forecasting system considering climate teleconnection factors according to claim 7, characterized in that, The process involves calculating the mutual information value of a standardized climate index for the historical resource sequence of the target site, using the lag time corresponding to the maximum mutual information value as the optimal early warning window, and constructing a forecast factor set, including: The maximum correlation time delay search algorithm based on mutual information is used to target the historical resource sequence of the site. Calculate its relationship with various climate factors Different lag times , Mutual information value under the moon : Select The largest value As the optimal early warning window for this factor at this station, a forecast factor set is constructed, in which... Let be a joint probability distribution, representing the value of the local resource. And the climate factor values ​​are The probability of them happening simultaneously This represents the marginal probability distribution of local resource data, that is, considering the probability of resource data occurrence in isolation. This represents the marginal probability distribution of climate factor data, i.e., the probability of a climate factor value occurring when considered individually.

11. The medium- and long-term wind and solar resource forecasting system considering climate teleconnection factors according to claim 7, characterized in that, The process of obtaining the current climate state vector based on the forecast factor set, searching the historical database for the historical year with the closest Euclidean distance to the current state, and extracting measured wind and solar resource data from the same period of the historical year to construct a baseline probability density function includes: Key factors selected based on the forecast factor set and its rate of change Constitutes the current climate state vector : in For the current moment, The total number of key climate factors selected; Search the historical database for the Euclidean algorithm that is closest to the current state. Historical years: Extract measured wind / solar resource data from the same period in similar years to construct a baseline probability density function (PDF) for prediction, where: For the current year (prediction starting point) of the [number]th State values ​​of each climate factor For the first The first historical year in the same period State values ​​of each climate factor For the current year's climate state and the first The Euclidean distance between historical years indicates that the climate modes of the two years are more similar. The total number of key climate factor characteristics involved in the calculation. An index for climate factor characteristics, from 1 to , For the first The weighting coefficients of each climate factor, reflecting the importance of that factor to local resources, are used to calculate the weighted similarity. For the current year, the State values ​​of each climate factor For the first The first historical year in the same period The state values ​​of each climate factor.

12. The medium- and long-term wind and solar resource forecasting system considering climate teleconnection factors according to claim 7, characterized in that, The process of using a baseline probability density function as a prior distribution and combining it with a preset dynamic mode prediction result as a likelihood function to output a prediction result containing deterministic values ​​and confidence intervals includes: A prediction equation is established, in which a nonlinear response term of climate factors is introduced. : in To predict resource quantity, This is the climatological mean. This is a non-linear mapping function fitted using machine learning. , These represent the proportions or contributions of climatological mean, nonlinear terms of climate factors, and dynamic model forecast terms in the final prediction results. This represents specific climate factor values ​​(such as the ENSO index), which here indicate the key climate predictors input into the model. These are values ​​predicted by seasonal dynamic models. It is a residual term or a random error term; The output includes deterministic values ​​and confidence intervals based on the distribution of similar historical years.

13. A computer device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the steps of the medium- and long-term wind and solar resource prediction method that considers climate teleconnection factors as described in any one of claims 1 to 6.

14. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by the processor, it implements the steps of the medium- and long-term wind and solar resource prediction method that considers climate teleconnection factors as described in any one of claims 1 to 6.

Images