New energy power station output scene generation system
By integrating geographical characteristics and meteorological data, a new energy power station output scenario model is constructed, and the model structure is optimized by the actual power generation, the accuracy and real-time problems of traditional prediction models are solved, and the power generation forecast and operation efficiency of new energy power stations are improved. .
Patent Information
- Application Number
- CN202411506100.1
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-10-28
- Publication Date
- 2024-11-26
- Estimated Expiration
- Not applicable · inactive patent
AI Technical Summary
The traditional new energy power generation prediction model cannot fully consider the complex relationships of multiple variables, resulting in inaccurate prediction results, and the data processing method depends on batch processing, which cannot achieve real-time updates and rapid feedback, affecting the timeliness of decision-making.
By collecting geographical characteristics, meteorological and historical power generation data, integrating these data and establishing relationships, building a output scenario model, and optimizing the model structure by comparing actual power generation with prediction results, dynamically adjusting the model to improve prediction accuracy and decision support.
It improves the accuracy and long-term prediction effect of power generation forecasting, supports timely decision-making, realizes dynamic response to actual situations and model adjustments, and improves the operational efficiency and prediction accuracy of new energy power plants.
Smart Images

Figure CN119026760A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the field of renewable energy power generation technology, and in particular to a renewable energy power station output scene generation system. Background Art
[0002] With the world's increasing attention to renewable energy, new energy power generation (such as solar energy, wind energy, etc.) has gradually become an important part of electricity supply. Traditional models may not be able to fully consider the complex relationship between multiple variables, resulting in inaccurate prediction results. Traditional data processing methods usually rely on batch processing and cannot achieve real-time data updates and rapid feedback, affecting the timeliness of decision-making.
[0003] Therefore, the present invention provides a new energy power station output scene generation system. Summary of the invention
[0004] A new energy power station output scenario generation system provided by the present invention collects geographical features, meteorological and historical power generation data, integrates these data and establishes relationships, builds an output scenario model based on the relationship results, and finally optimizes the model structure by comparing the actual power generation with the predicted results, thereby improving the accuracy of power generation prediction, supporting timely decision-making, dynamically adjusting the model according to actual conditions, and improving long-term prediction effects.
[0005] The present invention provides a new energy power station output scene generation system, comprising: Data collection module: collects geographical feature data of the new energy power station, determines meteorological data based on the geographical feature data, and obtains historical power generation data of the power station; Connection module: Integrate geographic feature data, meteorological data and historical power generation data into relevant data sets according to timestamps. At the same time, perform relational connection on each item in the relevant data sets to obtain relational connection results. Model generation module: constructs an output scenario model of a new energy power station based on the relationship connection results, generates corresponding output scenarios according to the output scenario model, and outputs a power generation time series; Adjustment module: compare the output power generation time series with the actual power generation data, and adjust and improve the model structure of the output scenario model according to the comparison results.
[0006] The present invention provides a new energy power station output scene generation system, a data acquisition module, including: Geographical feature data collection unit: performs feature analysis on the new energy power station to obtain the power station features, determines the geographical features of the new energy power station based on the power station features, and collects geographical feature data based on the geographical features; Meteorological data collection unit: Determine the meteorological collection focus based on the power station characteristics and geographical feature data of the new energy power station, and collect meteorological data according to the meteorological collection focus.
[0007] The present invention provides a new energy power station output scene generation system, a connection module, including: Alignment unit: unify the timestamp format of geographic feature data, meteorological data and historical power generation data, align the geographic feature data, meteorological data and historical power generation data according to the timestamp, and integrate them into relevant data sets; Analysis unit: Generate a connection buffer based on geographic feature data, spatially connect the meteorological data with the connection buffer, extract the meteorological data in the area within the connection buffer, calculate the statistical information of the meteorological data in the area, identify the change trend of the meteorological data in the area according to the statistical information, determine the power fluctuation of the historical power generation data in each connection buffer according to the change trend, and then determine the impact of the change trend on the power generation efficiency of the historical power generation data.
[0008] The present invention provides a new energy power station output scene generation system, a connection module, and further includes: Connection unit: Establish a first relational connection between geographic feature data and meteorological data based on the statistical information, establish a second relational connection between meteorological data and historical power generation data based on the power generation efficiency impact, and perform relational connection on each item in the relevant data set according to the first relational connection and the second relational connection to obtain a relational connection result.
[0009] The present invention provides a new energy power station output scenario generation system, a model generation module, including: Target classification unit: obtains the development target of the output scenario of the new energy power station, classifies the development target into types, and obtains long-term and short-term targets; Model building unit: respectively combining the long-term goal and the short-term goal with the relationship connection result, designing the first model framework and the second model framework, and at the same time, determining the first input variable and the first output variable of the output scenario model according to the long-term goal, determining the second input variable and the second output variable according to the short-term goal, determining the first scenario model by combining the first model framework, the first output variable and the first input variable, determining the second scenario model by combining the second model framework, the second output variable and the second input variable, and constructing the output scenario model of the new energy power station according to the first scenario model and the second scenario model; A generating unit generates an output scenario based on the output scenario model, and outputs a time series of power generation under a corresponding target.
[0010] The present invention provides a new energy power station output scenario generation system, a model building unit, comprising: Variable determination subunit: selecting long-term influencing factors as the first input variable according to the long-term goal from the goal-factor table, and selecting short-term influencing factors as the second input variable according to the short-term goal; The first model subunit: creates a hysteresis feature based on the first input variable, and constructs a first scenario model by combining the hysteresis feature with the first model framework; The second model subunit adds a time feature based on the second input variable, and constructs a second scenario model by combining the time feature with the second model framework.
[0011] The present invention provides a new energy power station output scene generation system, a generation unit, including: Condition subunit: defines the first meteorological condition according to the long-term goal and defines the second meteorological condition according to the short-term goal; State subunit: analyzing a first operating state of the new energy power station equipment based on a first to-be-realized period corresponding to a long-term goal and a first meteorological condition, and analyzing a second operating state of the new energy power station equipment based on a second to-be-realized period corresponding to a short-term goal and a second meteorological condition; Parameter subunit: defines long-term output scenario parameters based on long-term goals and the first meteorological condition, and defines short-term output scenario parameters based on short-term goals and the second meteorological condition; Output subunit: input the long-term output scenario parameters and the first operating state into the output scenario model, output the first power generation sequence of the long-term target, input the short-term output scenario parameters and the second operating state into the output scenario model, output the second power generation sequence of the short-term target, and combine the first power generation sequence and the second power generation sequence to obtain the power generation time series.
[0012] The present invention provides a new energy power station output scene generation system, an adjustment module, including: Fitting judgment unit: compare the actual power generation data with the power generation time series, obtain the comparison results and draw the residual image. At the same time, determine the baseline of the residual image based on the historical power generation data, and judge the model fitting situation according to the positional relationship between the baseline and the residual image; Image judgment unit: based on the model fitting situation, calculate the autocorrelation coefficient of the residual, obtain the autocorrelation coefficient and partial autocorrelation coefficient, draw the autocorrelation graph and partial autocorrelation graph, observe the autocorrelation coefficient from the autocorrelation graph, judge the sensitivity of the output scenario model in capturing the time series dependency of the power generation, obtain the sensitivity judgment result, analyze the lag period from the partial autocorrelation graph, and judge the order of the output scenario model; Structural adjustment unit: adjusts the model structure of the output scenario model according to the sensitivity judgment and the order judgment.
[0013] The present invention provides a new energy power station output scene generation system, an image judgment unit, comprising: Autocorrelation coefficient subunit: ,in, The residual at time point t; represents the residual at time point tk; Represents the residual mean; n represents the total number of observed data points for calculating the residual; represents the autocorrelation coefficient, which indicates the correlation between the residuals at time point t and time point tk; k represents the number of lags, which indicates the time interval; Partial autocorrelation coefficient subunit: , where cov() represents the covariance operation; represents the partial autocorrelation coefficient, which indicates the direct correlation between the residual at time point t and the residual at time point t−k after controlling for the influence of all lag terms below the kth order; represents the correlation between the residuals at time t and time t−j after controlling for the effects of all lagged terms below order j; j represents the number of lagged terms that have been controlled.
[0014] The present invention provides a new energy power station output scene generation system, a structure adjustment unit, comprising: Sensitivity adjustment subunit: if the sensitivity judgment result is higher than the preset sensitivity range, the point density of the power generation time series is increased; if the sensitivity judgment result is lower than the preset sensitivity range, the output scenario model is simplified; Order adjustment subunit: Determine the order of the output scenario model from the coefficient-order correspondence table according to the calculated autocorrelation coefficient and partial autocorrelation coefficient. If the order is higher than the preset order range, simplify the output scenario model. If the order is lower than the preset order range, increase the point density of the power generation time series.
[0015] Compared with the prior art, the beneficial effects of the present application are as follows: by collecting geographical features, meteorological and historical power generation data, integrating these data and establishing relationships, building an output scenario model based on the relationship results, and finally optimizing the model structure by comparing the actual power generation with the predicted results, the accuracy of power generation prediction is improved, timely decision-making is supported, the model is dynamically adjusted according to actual conditions, and the long-term prediction effect is improved.
[0016] Other features and advantages of the present invention will be described in the following description, and partly become apparent from the description, or understood by practicing the present invention. The purpose and other advantages of the present invention can be realized and obtained by the structures particularly pointed out in the written description and the accompanying drawings.
[0017] The technical solution of the present invention is further described in detail below through the accompanying drawings and embodiments. BRIEF DESCRIPTION OF THE DRAWINGS
[0018] The accompanying drawings are used to provide a further understanding of the present invention and constitute a part of the specification. Together with the embodiments of the present invention, they are used to explain the present invention and do not constitute a limitation of the present invention. In the accompanying drawings: Figure 1 It is a structural schematic diagram of a new energy power station output scenario generation system provided by an embodiment of the present invention. DETAILED DESCRIPTION
[0019] The preferred embodiments of the present invention are described below in conjunction with the accompanying drawings. It should be understood that the preferred embodiments described herein are only used to illustrate and explain the present invention, and are not used to limit the present invention.
[0020] Embodiment 1: The embodiment of the present invention provides a system for generating output scenarios of new energy power stations, such as Figure 1 As shown, including: Data collection module: collects geographical feature data of the new energy power station, determines meteorological data based on the geographical feature data, and obtains historical power generation data of the power station; Connection module: Integrate geographic feature data, meteorological data and historical power generation data into relevant data sets according to timestamps. At the same time, perform relational connection on each item in the relevant data sets to obtain relational connection results. Model generation module: constructs an output scenario model of a new energy power station based on the relationship connection results, generates corresponding output scenarios according to the output scenario model, and outputs a power generation time series; Adjustment module: compare the output power generation time series with the actual power generation data, and adjust and improve the model structure of the output scenario model according to the comparison results.
[0021] In this embodiment, geographical features refer to the natural and cultural features of the environment in which the power station is located, including topography, climate, land use, etc. For example, the altitude of the area where the power station is located (such as 500 meters), land use type (such as agricultural land, urban land), and geographical feature data refers to specific data information that describes geographical features, such as terrain data (such as digital elevation model) and land use map (such as land use classification data).
[0022] In this embodiment, meteorological data refers to specific data information related to meteorological conditions, such as sunshine intensity and wind speed data.
[0023] In this embodiment, historical power generation data refers to the quantitative record of actual power generation of the new energy power station in the past period of time. These data usually include: timestamp: the specific time corresponding to each power generation record, usually in hours, days or months, power generation value: the total amount of power generated by the power station in a specific time period, usually in kilowatt-hours (kWh) or megawatt-hours (MWh), data source: these data usually come from the power station's monitoring system or power grid company to ensure their accuracy and reliability.
[0024] In this embodiment, the relevant data set refers to a data set that is formed by integrating geographic feature data, meteorological data, and historical power generation data through timestamps in a connection module.
[0025] In this embodiment, the relationship connection result is obtained by integrating the first and second relationship connections, resulting in the mutual influence and connection between the various data. For example, in the generation of new energy power station output scenarios, the relationship connection result can be a power generation prediction model under different meteorological conditions, showing how the power generation capacity of the power station changes under the influence of specific geographical conditions and meteorological factors.
[0026] In this embodiment, the relationship connection process is to establish a relationship connection between geographic feature data and meteorological data by analyzing statistical information, and at the same time establish a relationship connection between meteorological data and historical power generation data based on the impact of power generation efficiency. These two relationship connections are applied to each item in the relevant data set, and a comprehensive analysis is performed to obtain the relationship connection results.
[0027] In this embodiment, the output scenario model integrates the first and second scenario models to form a complete power generation capacity prediction model, for example, an integrated model that can provide annual and daily power generation forecasts to support different decision-making needs.
[0028] In this embodiment, the power generation time series includes a first power generation series and a second power generation series, which is a data series of power generation changing over time generated based on the output scenario model, for example, power generation forecast data for each month and each day in the next 12 months, forming a power generation time series for analysis and decision support.
[0029] In this embodiment, by obtaining the long-term and short-term development goals of the output scenarios of the new energy power station and combining these goals with the relationship connection results, a corresponding model framework is constructed, and the input and output variables are determined according to different goals to form the first and second scenario models. Finally, the two models are combined to generate the output scenario and output the power generation time series.
[0030] In this embodiment, the comparison result is the difference between the actual power generation data and the predicted power generation. For example, assuming that the actual power generation of a solar power station on a certain day is 500 kWh and the predicted power generation is 450 kWh, the comparison result is a residual of 50 kWh.
[0031] In this embodiment, the specific process corresponding to the adjustment of the model structure is to determine whether to increase or decrease the lag term based on the sensitivity and order judgment results, rebuild the model, add new features (such as meteorological data, historical power generation, etc.) or adjust existing features, and perform model training and verification to ensure that the adjusted model has improved prediction capabilities. For example, if it is found that the model is less sensitive to changes in wind speed, wind speed can be added as an input feature and the model can be retrained to improve the prediction accuracy of power generation.
[0032] The working principle and beneficial effects of the above technical solution are: by collecting geographical features, meteorological and historical power generation data, integrating these data and establishing relationships, building an output scenario model based on the relationship results, and finally optimizing the model structure by comparing the actual power generation with the predicted results, the accuracy of power generation prediction is improved, timely decision-making is supported, the model is dynamically adjusted according to actual conditions, and the long-term prediction effect is improved.
[0033] Embodiment 2: The embodiment of the present invention provides a new energy power station output scenario generation system, a data acquisition module, including: Geographical feature data collection unit: performs feature analysis on the new energy power station to obtain the power station features, determines the geographical features of the new energy power station based on the power station features, and collects geographical feature data based on the geographical features; Meteorological data collection unit: Determine the meteorological collection focus based on the power station characteristics and geographical feature data of the new energy power station, and collect meteorological data according to the meteorological collection focus.
[0034] In this embodiment, the feature analysis process is to collect basic information of the new energy power station, such as power generation type (solar energy, wind energy, etc.), installed capacity, operating status, etc., and analyze its operating performance and potential influencing factors.
[0035] In this embodiment, the power station characteristics refer to the basic attributes and performance indicators related to the new energy power station, such as the installed capacity of the solar power station (such as 100 MW), the number of wind turbines and the capacity of each unit in the wind farm (such as 50 units, each with 2 MW).
[0036] In this embodiment, the focus of meteorological collection refers to key meteorological factors determined according to the characteristics of the power station and the geographical characteristics. For example, for a solar power station, the focus is on collecting sunshine hours and radiation intensity data, and for a wind farm, the focus is on collecting wind speed and wind direction data.
[0037] The working principle and beneficial effects of the above technical solution are: by analyzing the characteristics of new energy power stations, combining geographical characteristics, determining the focus of meteorological data collection, thereby achieving a comprehensive analysis of the power station operating environment, and optimizing the performance prediction and decision support of the power station through real-time collection and processing of geographical and meteorological data, thereby improving power generation efficiency and the accuracy of power generation prediction of new energy power stations.
[0038] Embodiment 3: The embodiment of the present invention provides a new energy power station output scenario generation system, a connection module, including: Alignment unit: unify the timestamp format of geographic feature data, meteorological data and historical power generation data, align the geographic feature data, meteorological data and historical power generation data according to the timestamp, and integrate them into relevant data sets; Analysis unit: Generate a connection buffer based on geographic feature data, spatially connect the meteorological data with the connection buffer, extract the meteorological data in the area within the connection buffer, calculate the statistical information of the meteorological data in the area, identify the change trend of the meteorological data in the area according to the statistical information, determine the power fluctuation of the historical power generation data in each connection buffer according to the change trend, and then determine the impact of the change trend on the power generation efficiency of the historical power generation data.
[0039] In this embodiment, the connection buffer refers to an area created around a specific geographical feature (such as the location of a power plant) for analyzing relevant data within the area. For example, a buffer with a radius of 5 kilometers is established around a solar power plant to analyze the meteorological and geographical features within the area.
[0040] In this embodiment, spatial connection refers to associating different geographic data sets (such as meteorological data and geographic feature data) based on geographic location to extract relevant information. For example, the wind speed data of the meteorological station is spatially connected with the connection buffer of the power station to obtain the wind speed information of the area.
[0041] In this embodiment, the intra-region meteorological data refers to meteorological information collected in a specific connection buffer zone, for example, temperature, humidity and precipitation data collected in the connection buffer zone of a power station.
[0042] In this embodiment, the statistical information is a numerical value obtained by summarizing and analyzing the meteorological data in the area, such as the mean, variance, etc. For example, the average monthly temperature and precipitation in the connection buffer zone in the past year are calculated to evaluate changes in meteorological conditions.
[0043] In this embodiment, the change trend refers to the long-term change direction or pattern of meteorological data identified through statistical information. For example, it is found that the average annual temperature in the connection buffer zone increases year by year, showing a trend of climate warming.
[0044] In this embodiment, the power fluctuation refers to the change of historical power generation data within a certain time range. For example, analysis shows that the power fluctuation in a certain quarter is large, which may be caused by changes in meteorological conditions during that period.
[0045] In this embodiment, the impact on power generation efficiency refers to the degree of influence of meteorological changes on the power generation capacity of the new energy power station. For example, through analysis, it is found that when the sunshine intensity in the connection buffer zone increases, the power generation efficiency of the power station increases by 15%.
[0046] The working principle and beneficial effects of the above technical solution are: by aligning geographic features, meteorological data and historical power generation data to form a unified data set, generating a connection buffer, and performing spatial connection and statistical analysis of meteorological data, so as to identify meteorological change trends and their impact on historical power generation and optimize the evaluation of power generation efficiency.
[0047] Embodiment 4: The embodiment of the present invention provides a new energy power station output scenario generation system, a connection module, and further includes: Connection unit: Establish a first relational connection between geographic feature data and meteorological data based on the statistical information, establish a second relational connection between meteorological data and historical power generation data based on the power generation efficiency impact, and perform relational connection on each item in the relevant data set according to the first relational connection and the second relational connection to obtain a relational connection result.
[0048] In this embodiment, the first relationship connection is an association between geographic feature data and meteorological data established based on statistical information. For example, in the generation of new energy power station output scenarios, the first relationship connection can be to analyze the relationship between the terrain features of a certain area (such as altitude, slope) and the meteorological data of the area, such as wind speed and sunshine duration, to determine the impact of geographical features on meteorological conditions.
[0049] In this embodiment, the second relationship connection is an association between meteorological data and historical power generation data based on the impact on power generation efficiency. For example, in output scenario generation, the second relationship connection can be an analysis of the impact of specific meteorological conditions (such as temperature, humidity, wind speed) on historical power generation, such as the changes in power generation of a photovoltaic power station within a specific temperature range.
[0050] The working principle and beneficial effects of the above technical solution are: by analyzing statistical information, a relational connection is established between geographic feature data and meteorological data, and at the same time, a relational connection is established between meteorological data and historical power generation data based on the impact of power generation efficiency. These two relational connections are applied to each item in the relevant data set, and a comprehensive analysis is performed to obtain the relational connection results to reveal the mutual influence between different factors, which helps to improve the power generation efficiency of new energy power stations.
[0051] Embodiment 5: The embodiment of the present invention provides a new energy power station output scenario generation system, a model generation module, including: Target classification unit: obtains the development target of the output scenario of the new energy power station, classifies the development target into types, and obtains long-term and short-term targets; Model building unit: respectively combining the long-term goal and the short-term goal with the relationship connection result, designing the first model framework and the second model framework, and at the same time, determining the first input variable and the first output variable of the output scenario model according to the long-term goal, determining the second input variable and the second output variable according to the short-term goal, determining the first scenario model by combining the first model framework, the first output variable and the first input variable, determining the second scenario model by combining the second model framework, the second output variable and the second input variable, and constructing the output scenario model of the new energy power station according to the first scenario model and the second scenario model; A generating unit generates an output scenario based on the output scenario model, and outputs a time series of power generation under a corresponding target.
[0052] In this embodiment, the development goal is the power generation capacity and operation target of the new energy power station in different time periods, such as improving power generation efficiency, reducing operating costs, and increasing the proportion of renewable energy.
[0053] In this embodiment, the long-term goal is the goal that you hope to achieve within a longer period of time (for example, 5-10 years), such as achieving a 30% increase in annual power generation, building new power generation facilities, and achieving carbon neutrality goals; the short-term goal is: the goal that you hope to achieve within a shorter period of time (for example, within 1 year), such as improving daily power generation efficiency by 5%, optimizing maintenance plans to reduce downtime, and improving customer satisfaction.
[0054] In this embodiment, the first model framework is a model structure for long-term goals, for example, based on historical data and meteorological models, a long-term power generation forecasting model is constructed, and a machine learning algorithm is used to analyze the power generation potential in different seasons.
[0055] In this embodiment, the second model framework is a model structure for short-term goals, such as a real-time data monitoring model that combines weather forecasts and grid load forecasts to perform short-term power generation scheduling.
[0056] In this embodiment, the first output variable is the prediction result related to the long-term goal, such as the annual total power generation forecast; the first input variable is the factor affecting the long-term goal, such as historical power generation data, climate change trends, and equipment performance indicators.
[0057] In this embodiment, the second input variable is a factor that affects the short-term goal, such as real-time meteorological data (temperature, wind speed), grid load demand, and equipment status information; the second output variable is a prediction result related to the short-term goal, such as daily power generation prediction.
[0058] In this embodiment, the first scenario model is a model constructed based on long-term goals and input variables, such as an annual power generation prediction model generated by analyzing historical data, which outputs power generation capacity in different seasons; the second scenario model is a model constructed based on short-term goals and input variables, such as a real-time power generation scheduling model, which predicts power generation in the next few hours based on current weather and grid demand.
[0059] The working principle and beneficial effects of the above technical solution are: by obtaining the long-term and short-term development goals of the output scenarios of new energy power stations, and combining these goals with the relationship connection results, a corresponding model framework is constructed, and the input and output variables are determined according to different goals to form the first and second scenario models. Finally, the two models are combined to generate output scenarios and output power generation time series to improve the accuracy and real-time performance of power generation forecasts.
[0060] Embodiment 6: The embodiment of the present invention provides a new energy power station output scenario generation system, a model building unit, including: Variable determination subunit: selecting long-term influencing factors as the first input variable according to the long-term goal from the goal-factor table, and selecting short-term influencing factors as the second input variable according to the short-term goal; The first model subunit: creates a hysteresis feature based on the first input variable, and constructs a first scenario model by combining the hysteresis feature with the first model framework; The second model subunit adds a time feature based on the second input variable, and constructs a second scenario model by combining the time feature with the second model framework.
[0061] In this embodiment, the target-factor table is a systematic table that lists the targets related to the output scenarios of the new energy power station and their corresponding influencing factors. For example, the table may include long-term targets (such as annual power generation) and their influencing factors (such as meteorological conditions and equipment performance), as well as short-term targets (such as daily power generation) and their influencing factors (such as real-time weather data and load demand).
[0062] In this embodiment, long-term influencing factors are factors that affect the realization of long-term goals, and usually involve changes over a longer period of time, such as climate change trends, equipment aging, policies and regulations, and changes in market demand.
[0063] In this embodiment, short-term influencing factors are factors that affect the achievement of short-term goals, and usually involve changes in a relatively short period of time, such as real-time weather conditions (such as wind speed, temperature), grid load, and equipment status (such as maintenance status).
[0064] In this embodiment, the hysteresis feature is a feature generated based on historical data, reflecting the state or behavior within a certain period of time in the past, such as the power generation in the past 7 days and the average temperature in the past 30 days.
[0065] In this embodiment, the time feature is a feature related to time and is used to represent the time dimension of data, such as hours, days of the week, seasons, months, etc.
[0066] In this embodiment, the process of constructing the first scenario model is to select long-term input variables: select long-term influencing factors related to long-term goals from the target-factor table; based on the selected long-term input variables, create lagged features to capture the impact of historical data; combine the lagged features with the first model framework, use appropriate algorithms (such as regression analysis, machine learning, etc.) to construct the first scenario model, use historical data to train the model, and evaluate the model performance through a validation set.
[0067] In this embodiment, the process of constructing the second scenario model is to select short-term influencing factors related to short-term goals from the target-factor table, add time features based on the selected short-term input variables to reflect the time dimension of the data, combine the time features with the second model framework, use appropriate algorithms to construct the second scenario model, train the model using real-time data or short-term historical data, and evaluate the model performance through a validation set.
[0068] The working principle and beneficial effects of the above technical solution are: by selecting long-term and short-term influencing factors from the target-factor table as the first input variable and the second input variable respectively, and combining the lag characteristics with the first model framework, a long-term power generation scenario model is constructed, and by adding time characteristics and combining with the second model framework, a short-term power generation scenario model is constructed, thereby achieving accurate power generation prediction. At the same time, clear input variables can help analyze the impact of different factors on power generation and enhance the interpretability of the model.
[0069] Embodiment 7: An embodiment of the present invention provides a new energy power station output scenario generation system, a generation unit, including: Condition subunit: defines the first meteorological condition according to the long-term goal and defines the second meteorological condition according to the short-term goal; State subunit: analyzing a first operating state of the new energy power station equipment based on a first to-be-realized period corresponding to a long-term goal and a first meteorological condition, and analyzing a second operating state of the new energy power station equipment based on a second to-be-realized period corresponding to a short-term goal and a second meteorological condition; Parameter subunit: defines long-term output scenario parameters based on long-term goals and the first meteorological condition, and defines short-term output scenario parameters based on short-term goals and the second meteorological condition; Output subunit: input the long-term output scenario parameters and the first operating state into the output scenario model, output the first power generation sequence of the long-term target, input the short-term output scenario parameters and the second operating state into the output scenario model, output the second power generation sequence of the short-term target, and combine the first power generation sequence and the second power generation sequence to obtain the power generation time series.
[0070] In this embodiment, the first meteorological condition is a meteorological factor related to the long-term goal, such as the average annual temperature, precipitation, wind speed, etc. For example, the average annual temperature in a certain area is 20°C, the annual precipitation is 800 mm, and the average annual wind speed is 4 m / s.
[0071] In this embodiment, the second meteorological condition is a meteorological factor related to the short-term goal, such as the temperature, wind speed, humidity, etc. of the day or week. For example, the weather forecast for the next week shows: October 1: temperature 22°C, wind speed 5 m / s, cloud coverage 30%.
[0072] In this embodiment, the first period to be achieved is a time period related to the long-term goal, usually several months to a year. For example, January 2024 to December 2024 is selected as the first period to be achieved.
[0073] In this embodiment, the first operating state is the operating status of the new energy power station equipment during the first cycle to be implemented, including equipment health status, efficiency and failure rate, etc. For example, in 2024, the equipment health monitoring of the power station shows that the average power generation efficiency is 90% and the failure rate is 2%.
[0074] In this embodiment, the second period to be achieved is a time period related to the short-term goal, usually a few days to a few weeks. For example, October 1, 2024 to October 7, 2024 is selected as the second period to be achieved.
[0075] In this embodiment, the second operating state is the operating status of the new energy power station equipment during the second cycle to be implemented, which may include real-time monitoring data. For example, the monitoring data from October 1 to October 7, 2024 shows that the power generation efficiency is 85% and the load condition is 80%.
[0076] In this embodiment, the long-term output scenario parameters are parameters defined based on the first meteorological condition and the long-term goal, and are used to evaluate the long-term power generation capacity. For example, based on the first meteorological condition, the long-term output scenario parameters are set to an expected annual power generation of 500,000 kWh and an equipment utilization rate of 75%.
[0077] In this embodiment, the short-term output scenario parameters are parameters defined based on the second meteorological condition and the short-term goal, and are used to evaluate the short-term power generation capacity. For example, based on the second meteorological condition, the short-term output scenario parameters are set to an expected power generation of 10,000 kWh in the next week, and the equipment scheduling is adjusted on demand.
[0078] In this embodiment, the first power generation sequence is a power generation data sequence generated based on the long-term output scenario parameters and the first operating state. For example, the monthly power generation sequence is: January: 40,000 kWh, February: 35,000 kWh, March: 42,000 kWh, ..., December: 50,000 kWh.
[0079] In this embodiment, the second power generation sequence is a power generation data sequence generated based on the short-term output scenario parameters and the second operating state. For example, the power generation sequence in October 2024 is: October 1: 1,500 kWh, October 2: 1,200 kWh, October 3: 1,800 kWh,..., October 7: 1,600 kWh.
[0080] In this embodiment, how to obtain the power generation time series is to integrate the first power generation sequence and the second power generation sequence to obtain the overall power generation time series. For example, the first power generation sequence (such as annual power generation) is combined with the second power generation sequence (such as weekly power generation). The power generation time series that may be obtained is: January 2024 (40,000 kWh), February 2024 (35,000 kWh), March 2024 (42,000 kWh), October 1 (1,500 kWh), October 2 (1,200kWh), ..., October 7 (1,600 kWh).
[0081] The working principle and beneficial effects of the above technical solution are: by defining meteorological conditions, analyzing the operating status of equipment, formulating long-term and short-term output scenario parameters, inputting these parameters and the corresponding operating status into the output scenario model, generating long-term and short-term power generation series respectively, and finally comprehensively deriving the power generation time series to achieve accurate output scenario prediction, so that the power station can detect potential problems in a timely manner and improve the reliability of equipment.
[0082] Embodiment 8: The embodiment of the present invention provides a new energy power station output scenario generation system, an adjustment module, including: Fitting judgment unit: compare the actual power generation data with the power generation time series, obtain the comparison results and draw the residual image. At the same time, determine the baseline of the residual image based on the historical power generation data, and judge the model fitting situation according to the positional relationship between the baseline and the residual image; Image judgment unit: based on the model fitting situation, calculate the autocorrelation coefficient of the residual, obtain the autocorrelation coefficient and partial autocorrelation coefficient, draw the autocorrelation graph and partial autocorrelation graph, observe the autocorrelation coefficient from the autocorrelation graph, judge the sensitivity of the output scenario model in capturing the time series dependency of the power generation, obtain the sensitivity judgment result, analyze the lag period from the partial autocorrelation graph, and judge the order of the output scenario model; Structural adjustment unit: adjusts the model structure of the output scenario model according to the sensitivity judgment and the order judgment.
[0083] In this embodiment, the residual image refers to a graph formed by the difference (ie, residual) between the actual power generation data and the power generation time series predicted by the model during the model fitting process.
[0084] In this embodiment, the baseline is a reference line of the residual image, which is usually the mean or median of the residuals. For example, if the mean residual value in the past week is 10 kWh, the baseline is 10 kWh. This line is used to determine the fluctuation of the residuals.
[0085] In this embodiment, the model fit is the closeness between the model prediction results and the actual data. For example, if most of the residuals are within the range of ±20 kWh, it means that the model fits well; if the residuals fluctuate greatly, it may indicate that the model has deviations.
[0086] In this embodiment, the positional relationship is the relative position between the residual image and the baseline. For example, if most of the residual image is above the baseline, it means that the model may underestimate the power generation; otherwise, it may overestimate it.
[0087] In this embodiment, the autocorrelation coefficient is a measure of the correlation between the power generation time series itself at different time lags. For example, if the autocorrelation coefficient is 0.8 at a lag of 1 hour, it means that the power generation in the current hour is highly correlated with the power generation in the previous hour.
[0088] In this embodiment, the partial autocorrelation coefficient is the correlation between the current value and the specific lagged value while controlling the influence of other lagged terms. For example, if the partial autocorrelation coefficient is significantly not zero after a lag of 2 hours, it means that after controlling other factors, the current power generation is still correlated with the power generation 2 hours ago.
[0089] In this embodiment, the lag period is the time interval between a data point in the power generation time series and its previous data point. For example, in a wind farm, a change in wind speed may take 2 hours to affect power generation, so the lag period is 2 hours.
[0090] In this embodiment, the order is the number of lag terms in the model. For example, if it is found through analysis that the model needs to consider the power generation of the previous 3 hours to predict the current power generation, the order of the model is 3.
[0091] In this embodiment, the sensitivity of the power generation time series dependency refers to the responsiveness of the power scenario model to changes in input data.
[0092] In this embodiment, the process of judging the sensitivity of the output scenario model in capturing the time series dependency of power generation is to calculate the autocorrelation and partial autocorrelation coefficients of the residual, observe the significance of the autocorrelation coefficient, and judge the sensitivity of the model to the time series dependency of power generation. For example, if the autocorrelation coefficient is significantly not zero after a lag of 1 hour and 2 hours, it means that the model can sensitively capture short-term dependencies.
[0093] In this embodiment, the process of determining the order of the output scenario model is to observe the partial autocorrelation coefficient, find the last significant partial autocorrelation coefficient, and determine the order of the model. For example, if the partial autocorrelation coefficient becomes insignificant after a lag of 3 hours, it means that the order of the model should be 3.
[0094] In this embodiment, the sensitivity judgment condition refers to the model's ability to respond to changes in input data. For example, if the power generation prediction of the model changes significantly when the input wind speed changes, it means that the sensitivity is high and it is suitable for real-time prediction.
[0095] In this embodiment, the order determination is the number of model lag terms obtained through autocorrelation and partial autocorrelation analysis. For example, if the analysis results show that the lags of 1 hour and 2 hours are significant, and the lag of 3 hours is not significant, then the order of the model is 2.
[0096] The working principle and beneficial effects of the above technical solution are: by comparing the actual power generation data with the predicted power generation time series, generating a residual image and determining a baseline to evaluate the model fitting effect, calculating the autocorrelation and partial autocorrelation coefficients of the residual, analyzing the power generation time series dependency and model order, judging the results based on sensitivity and order, optimizing the model structure, thereby improving the prediction accuracy, enhancing the adaptability to complex variable relationships, making power generation prediction more accurate, reducing the uncertainty of power supply, and enhancing the stability and reliability of the power system.
[0097] Embodiment 9: The embodiment of the present invention provides a new energy power station output scene generation system, an image judgment unit, including: Autocorrelation coefficient subunit: ,in, The residual at time point t; represents the residual at time point tk; Represents the residual mean; n represents the total number of observed data points for calculating the residual; represents the autocorrelation coefficient, which indicates the correlation between the residuals at time point t and time point tk; k represents the number of lags, which indicates the time interval; Partial autocorrelation coefficient subunit: , where cov() represents the covariance operation; represents the partial autocorrelation coefficient, which indicates the direct correlation between the residual at time point t and the residual at time point t−k after controlling for the influence of all lag terms below the kth order; represents the correlation between the residuals at time t and time t−j after controlling for the effects of all lagged terms below order j; j represents the number of lagged terms that have been controlled.
[0098] In this embodiment, k is the interval between the current time point and the past time point. For example, if k=1, it indicates the autocorrelation between the current time point and the previous time point; if k=2, it indicates the autocorrelation between the previous two time points, and so on.
[0099] In this embodiment, j is the number of other lags that have been controlled (or eliminated) in the partial autocorrelation calculation, for example, in the calculation When , the partial autocorrelations of the first k − 1 lags may be used, and these lags are denoted by j.
[0100] The working principle and beneficial effects of the above technical solution are: by measuring the correlation between the residual at time point t and the residual at lag k, the model's ability to capture the time series dependency of power generation is evaluated; the partial autocorrelation coefficient quantifies the direct relationship between the residual at time point t and the residual at lag k after controlling the influence of low-order lag terms, thereby providing a more accurate judgment of the model order; by improving the model to fully consider the complex relationship between multiple variables, the prediction accuracy can be improved, real-time data updates and rapid feedback can be supported, thereby enhancing the timeliness and effectiveness of decision-making, and optimizing power dispatch and resource allocation.
[0101] Embodiment 10: The embodiment of the present invention provides a new energy power station output scenario generation system, a structure adjustment unit, including: Sensitivity adjustment subunit: if the sensitivity judgment result is higher than the preset sensitivity range, the point density of the power generation time series is increased; if the sensitivity judgment result is lower than the preset sensitivity range, the output scenario model is simplified; Order adjustment subunit: Determine the order of the output scenario model from the coefficient-order correspondence table according to the calculated autocorrelation coefficient and partial autocorrelation coefficient. If the order is higher than the preset order range, simplify the output scenario model. If the order is lower than the preset order range, increase the point density of the power generation time series.
[0102] In this embodiment, the preset sensitivity range refers to a sensitivity threshold set during the model design phase, which is used to evaluate the responsiveness of the model to changes in input variables.
[0103] In this embodiment, simplification refers to reducing the complexity of the model, which is usually achieved by reducing the number of variables, lowering the model order, or using a simpler algorithm.
[0104] In this embodiment, the sampling point density refers to the number of data points collected per unit time in the power generation time series data. Increasing the sampling point density means collecting more observations in a specific time period.
[0105] In this embodiment, the coefficient-order correspondence table is a tool for mapping autocorrelation coefficients and partial autocorrelation coefficients to model orders. The table is usually based on empirical rules or statistical theories and provides optimal model orders corresponding to different autocorrelation and partial autocorrelation values.
[0106] The working principle and beneficial effects of the above technical solution are: based on sensitivity analysis and order judgment, the point density and model complexity of the power generation time series are dynamically adjusted. When the sensitivity is higher than the preset range, the point density is increased to capture more information; otherwise, the model is simplified to improve efficiency. The order judgment determines the model order through the autocorrelation and partial autocorrelation coefficients. When it exceeds the range, the model is simplified, and when it is lower than the range, the point density is increased to ensure the adaptability and accuracy of the model, optimize the calculation efficiency, reduce the risk of overfitting, and thus enhance the reliability and practicality of the output scenario model.
[0107] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.
Claims
1. A new energy power station output scenario generation system, characterized in that: include: Data collection module: collects geographical feature data of the new energy power station, determines meteorological data based on the geographical feature data, and obtains historical power generation data of the new energy power station; Connection module: Integrate geographic feature data, meteorological data and historical power generation data into relevant data sets according to timestamps. At the same time, perform relational connection on each item in the relevant data sets to obtain relational connection results. Model generation module: constructs an output scenario model of a new energy power station based on the relationship connection results, generates corresponding output scenarios according to the output scenario model, and outputs a power generation time series; Adjustment module: compare the output power generation time series with the actual power generation data, and adjust and improve the model structure of the output scenario model according to the comparison results.
2. A new energy power station output scenario generation system according to claim 1, characterized in that: Data acquisition module, including: Geographical feature data collection unit: performs feature analysis on the new energy power station to obtain the power station features, determines the geographical features of the new energy power station based on the power station features, and collects geographical feature data based on the geographical features; Meteorological data collection unit: Determine the meteorological collection focus based on the power station characteristics and geographical feature data of the new energy power station, and collect meteorological data according to the meteorological collection focus.
3. A new energy power station output scenario generation system according to claim 1, characterized in that: Connectivity module, including: Alignment unit: unify the timestamp format of geographic feature data, meteorological data and historical power generation data, align the geographic feature data, meteorological data and historical power generation data according to the timestamp, and integrate them into relevant data sets; Analysis unit: Generate a connection buffer based on geographic feature data, spatially connect the meteorological data with the connection buffer, extract the meteorological data in the area within the connection buffer, calculate the statistical information of the meteorological data in the area, identify the change trend of the meteorological data in the area according to the statistical information, determine the power fluctuation of the historical power generation data in each connection buffer according to the change trend, and then determine the impact of the change trend on the power generation efficiency of the historical power generation data.
4. A new energy power station output scenario generation system according to claim 3, characterized in that: The connection module also includes: Connection unit: Establish a first relational connection between geographic feature data and meteorological data based on the statistical information, establish a second relational connection between meteorological data and historical power generation data based on the power generation efficiency impact, and perform relational connection on each item in the relevant data set according to the first relational connection and the second relational connection to obtain a relational connection result.
5. A new energy power station output scenario generation system according to claim 1, characterized in that: Model generation module, including: Target classification unit: obtains the development target of the output scenario of the new energy power station, classifies the development target into types, and obtains long-term and short-term targets; Model building unit: respectively combining the long-term goal and the short-term goal with the relationship connection result, designing the first model framework and the second model framework, and at the same time, determining the first input variable and the first output variable of the output scenario model according to the long-term goal, determining the second input variable and the second output variable according to the short-term goal, determining the first scenario model by combining the first model framework, the first output variable and the first input variable, determining the second scenario model by combining the second model framework, the second output variable and the second input variable, and constructing the output scenario model of the new energy power station according to the first scenario model and the second scenario model; A generating unit generates an output scenario based on the output scenario model, and outputs a time series of power generation under a corresponding target.
6. A new energy power station output scenario generation system according to claim 5, characterized in that: Model building unit, including: Variable determination subunit: selecting long-term influencing factors as the first input variable according to the long-term goal from the goal-factor table, and selecting short-term influencing factors as the second input variable according to the short-term goal; The first model subunit: creates a hysteresis feature based on the first input variable, and constructs a first scenario model by combining the hysteresis feature with the first model framework; The second model subunit adds a time feature based on the second input variable, and constructs a second scenario model by combining the time feature with the second model framework.
7. A new energy power station output scenario generation system according to claim 5, characterized in that: Generate a unit, including: Condition subunit: defines the first meteorological condition according to the long-term goal and defines the second meteorological condition according to the short-term goal; State subunit: analyzing a first operating state of the new energy power station equipment based on a first to-be-realized period corresponding to a long-term goal and a first meteorological condition, and analyzing a second operating state of the new energy power station equipment based on a second to-be-realized period corresponding to a short-term goal and a second meteorological condition; Parameter subunit: defines long-term output scenario parameters based on long-term goals and the first meteorological condition, and defines short-term output scenario parameters based on short-term goals and the second meteorological condition; Output subunit: input the long-term output scenario parameters and the first operating state into the output scenario model, output the first power generation sequence of the long-term target, input the short-term output scenario parameters and the second operating state into the output scenario model, output the second power generation sequence of the short-term target, and combine the first power generation sequence and the second power generation sequence to obtain the power generation time series.
8. A new energy power station output scenario generation system according to claim 1, characterized in that: Adjustment modules, including: Fitting judgment unit: compare the actual power generation data with the power generation time series, obtain the comparison results and draw the residual image. At the same time, determine the baseline of the residual image based on the historical power generation data, and judge the model fitting situation according to the positional relationship between the baseline and the residual image; Image judgment unit: based on the model fitting situation, calculate the autocorrelation coefficient of the residual, obtain the autocorrelation coefficient and partial autocorrelation coefficient, draw the autocorrelation graph and partial autocorrelation graph, observe the autocorrelation coefficient from the autocorrelation graph, judge the sensitivity of the output scenario model in capturing the time series dependency of the power generation, obtain the sensitivity judgment result, analyze the lag period from the partial autocorrelation graph, and judge the order of the output scenario model; Structural adjustment unit: adjusts the model structure of the output scenario model according to the sensitivity judgment and the order judgment.
9. A new energy power station output scenario generation system according to claim 8, characterized in that: The image judgment unit comprises: Autocorrelation coefficient subunit: ,in, represents the residual at time point t; represents the residual at time point tk; Represents the residual mean; n represents the total number of observed data points for calculating the residual; represents the autocorrelation coefficient, which indicates the correlation between the residuals at time point t and time point tk; k represents the number of lags, which indicates the time interval; Partial autocorrelation coefficient subunit: , where cov() represents the covariance operation; represents the partial autocorrelation coefficient, which indicates the direct correlation between the residual at time point t and the residual at time point t−k after controlling for the influence of all lag terms below the kth order; represents the correlation between the residuals at time t and time t−j after controlling for the effects of all lagged terms below order j; j represents the number of lagged terms that have been controlled.
10. A new energy power station output scenario generation system according to claim 8, characterized in that: Structural adjustment unit, including: Sensitivity adjustment subunit: if the sensitivity judgment result is higher than the preset sensitivity range, the point density of the power generation time series is increased; if the sensitivity judgment result is lower than the preset sensitivity range, the output scenario model is simplified; Order adjustment subunit: Determine the order of the output scenario model from the coefficient-order correspondence table according to the calculated autocorrelation coefficient and partial autocorrelation coefficient. If the order is higher than the preset order range, simplify the output scenario model. If the order is lower than the preset order range, increase the point density of the power generation time series.
Citation Information
Patent Citations
Method for predicting economic output through electricity consumption data
CN110070202A
Construction method of combined model for accurately predicting deformation effect of concrete dam
CN111191191A
Auxiliary decision-making method for power generation and networking of new energy power station
CN114386703A
Medium and long term generating capacity prediction method and system based on new energy power generation
CN116865236A
Photovoltaic power generation power probability prediction method, system, chip and equipment
CN117318028A
