New energy station short-term power prediction method considering power prediction deviation

Abstract

The invention discloses a new energy station short-term power prediction method considering power prediction deviation. The method comprises the following steps: acquiring real-time meteorological data of a new energy station; based on the short-term power prediction model of the new energy station, combining the real-time meteorological data to calculate short-term power generation power prediction values of the new energy station in a plurality of subsequent detection periods; calibrating the short-term generated power predicted value based on a power prediction calibration model of the new energy station to obtain a short-term generated power prediction calibration value of the new energy station; wherein the calibration condition of the power prediction calibration model comprises a season factor and a day period factor. By setting calibration conditions of multiple dimensions and considering factors of seasons and time periods of days, meteorological conditions and power consumption mode changes can be coped with, the precision of short-term power prediction is improved, and prediction deviation is reduced; and the adaptability of short-term power prediction is enhanced, and the characteristics of power output due to different time and environment changes are met, so that the predicted value is more practical.

Description

Technical Field

[0001] The present invention relates to the field of new energy power generation prediction, and in particular to a new energy station short-term power prediction method taking into account power prediction deviation. Background Art

[0002] From the perspective of stable operation of the power system, the power output of new energy stations (such as wind power and photovoltaics) is intermittent and volatile. Short-term power forecasts can help the grid dispatching department understand the changes in new energy power in advance, so as to reasonably arrange conventional energy power generation plans, effectively balance power supply and demand, and ensure the frequency and voltage stability of the power grid. For power market transactions, accurate short-term power forecasts can provide new energy station operators with power expectation information, enabling them to participate in quotation and power allocation more reasonably in the power trading market and improve economic benefits. In addition, from the perspective of optimizing the operation of new energy stations themselves, the forecast results are helpful to adjust equipment operating parameters and arrange equipment maintenance plans in advance to cope with different power output scenarios and improve the safety and efficiency of station operations.

[0003] However, the construction location of new energy stations is relatively remote, and the meteorological conditions of the environment in which they are located vary significantly. Therefore, the short-term power generation of new energy stations is obviously affected by the changes in meteorological conditions when predicting. For example, there may be errors in meteorological observation equipment, and the accuracy and real-time performance of meteorological data are poor. In addition, the local meteorology is complex, and the errors of observation equipment and micro-meteorological phenomena affect the data quality. In terms of equipment characteristics and operating status, equipment performance changes dynamically, such as wind turbine blade wear, photovoltaic panel aging, and equipment failures that are difficult to predict, all of which will interfere with power output. The prediction model has limitations, is not adaptable enough to complex factors, and it is difficult to consider the interaction of multiple factors. In addition, the model parameters are plagued by timeliness and accuracy issues. Summary of the invention

[0004] The purpose of an embodiment of the present invention is to provide a short-term power prediction method for a new energy station taking into account power prediction deviations. By setting calibration conditions in multiple dimensions and taking into account seasonal and daily time factors, it can cope with changes in meteorological conditions and power consumption patterns, thereby improving the accuracy of short-term power predictions and reducing prediction deviations caused by these factors; it also enhances the adaptability of short-term power predictions to meet the characteristics of new energy station power output changes due to different times and environments, making the predicted values ​​more in line with reality.

[0005] In order to solve the above technical problems, a first aspect of an embodiment of the present invention provides a new energy station short-term power prediction method considering power prediction deviation, comprising the following steps:

[0006] Obtain real-time meteorological data of new energy stations;

[0007] Based on the short-term power prediction model of the new energy station, the short-term power prediction value of the new energy station in the subsequent several detection cycles is calculated in combination with the real-time meteorological data;

[0008] Calibrate the short-term power generation prediction value based on the power prediction calibration model of the new energy station to obtain the short-term power generation prediction calibration value of the new energy station;

[0009] The calibration conditions of the power prediction calibration model include: seasonal factors and daily time factors.

[0010] Furthermore, the short-term power prediction model based on the new energy station, combined with the real-time meteorological data, calculates the short-term power prediction value of the new energy station in the subsequent several detection cycles, including:

[0011] Obtain historical meteorological data and corresponding historical power generation data of new energy stations;

[0012] By means of the short-term power prediction model, a corresponding relationship between the historical meteorological data and the historical power generation data is established;

[0013] Based on the corresponding relationship and in combination with the real-time meteorological data, the short-term power generation forecast value of the new energy station is calculated.

[0014] Furthermore, the new energy station includes: a wind turbine generator set and a photovoltaic generator set;

[0015] The historical meteorological data include: the wind speed, wind direction, temperature and air pressure corresponding to the wind turbine generator set and the solar irradiance and temperature corresponding to the photovoltaic generator set.

[0016] Furthermore, the calculation formula of the short-term power generation prediction value of the wind turbine in the short-term power prediction model is:

[0017]

[0018] Among them, P w-p is the short-term power generation prediction value of the wind turbine, ρ is the air density value, A is the swept area of ​​the wind turbine blade, C P is the power coefficient, which is related to the tip speed ratio λ, pitch angle β, wind speed v, and actual temperature T w , actual air pressure p, wind direction is related to the angle θ of the fan main shaft, Δv is the wind speed loss, ΔP turbulence is the power loss caused by turbulence, f yaw (θ) is the wind direction correction function, f control (v,ω) is the fan control strategy function, which is related to the actual wind speed v and the impeller speed ω, f healthis the health status factor of the wind turbine, θ is the angle between the actual wind direction and the main axis direction of the wind turbine, v ci is the wind turbine cut-in wind speed, v r is the rated wind speed of the wind turbine, P wr is the rated power of the wind turbine.

[0019] Furthermore, the calculation formula of the short-term power generation prediction value of the photovoltaic power generation group in the short-term power prediction model is:

[0020]

[0021] Among them, P l-p is the short-term power generation prediction value of the photovoltaic generator set, η(T l ,I,α,β,γ) is the photoelectric conversion efficiency, α is the installation angle of the solar panel, β is the azimuth angle, γ is the solar altitude angle, T l is the actual temperature of the photovoltaic panels of the photovoltaic generator set, I is the actual solar irradiance value, S is the total area of ​​the photovoltaic panels of the photovoltaic generator set, T ref is the preset reference temperature, T co is the temperature coefficient, f shadow (x,y) is the shadow occlusion function, x and y are the distance and height of the occluder respectively, f dirt (d) is the dust coverage function, d is the dust thickness on the panel surface, f aging (t) is the aging function of the solar panel, and t is the service life of the solar panel.

[0022] Further, the power prediction calibration model based on the new energy station is used to calibrate the short-term power generation prediction value to obtain the short-term power generation prediction calibration value of the new energy station, including:

[0023] Obtaining historical power forecast data and corresponding historical power real data of the new energy station;

[0024] Based on the power prediction calibration model, calculating the average value of the difference between the historical power prediction data and the corresponding historical power real data;

[0025] When the difference average value is positive, the difference between the short-term power generation prediction value and the difference average value is calculated to obtain the short-term power generation prediction calibration value;

[0026] When the difference average value is negative, the sum of the short-term power generation prediction value and the absolute value of the difference average value is calculated to obtain the short-term power generation prediction calibration value.

[0027] Furthermore, after obtaining the short-term power generation prediction calibration value, the method further includes:

[0028] Calculate the difference average of the corresponding dimension according to the seasonal dimension and the daily time period dimension, obtain the change rule of the difference average of the corresponding dimension and the overall difference average, and obtain the adjustment coefficient of the difference average of the corresponding dimension and the overall difference average;

[0029] The short-term power prediction value is adjusted according to the adjustment coefficient to improve the accuracy of the short-term power prediction value.

[0030] Furthermore, the short-term power prediction method for new energy stations considering power prediction deviation also includes:

[0031] Respectively obtain the adjustment coefficients of the season dimension and the day time dimension;

[0032] The short-term power prediction values ​​of the wind turbine generator set and the photovoltaic generator set are calibrated based on the adjustment coefficient.

[0033] Furthermore, the calculation formula of the short-term power generation prediction calibration value of the wind turbine in the power prediction calibration model is:

[0034]

[0035] Among them, K season-i is the seasonal dimension adjustment coefficient, i is the seasonal number, K daytime-k is the daily time period dimension adjustment coefficient, and k is the corresponding serial number of the daily time period.

[0036] Furthermore, the calculation formula of the short-term power generation prediction calibration value of the photovoltaic power generation group in the short-term power prediction model is:

[0037]

[0038] Among them, K season-i is the seasonal dimension adjustment coefficient, i is the seasonal number, K daytime-k is the daily time period dimension adjustment coefficient, and k is the corresponding serial number of the daily time period.

[0039] Accordingly, a second aspect of an embodiment of the present invention provides a short-term power prediction device for a new energy station taking into account power prediction deviation, comprising:

[0040] A meteorological data acquisition module, which is used to obtain real-time meteorological data of new energy stations;

[0041] A power generation prediction module, which is used to calculate the short-term power generation prediction value of the new energy station in the subsequent several detection cycles based on the short-term power prediction model of the new energy station and in combination with the real-time meteorological data;

[0042] A power generation calibration module, which is used to calibrate the short-term power generation prediction value based on the power prediction calibration model of the new energy station to obtain the short-term power generation prediction calibration value of the new energy station;

[0043] The calibration conditions of the power prediction calibration model include: seasonal factors and daily time factors.

[0044] Accordingly, a third aspect of an embodiment of the present invention provides an electronic device, comprising: at least one processor; and a memory connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor executes the above-mentioned short-term power prediction method for new energy stations taking into account power prediction deviations.

[0045] Accordingly, a fourth aspect of an embodiment of the present invention provides a computer-readable storage medium having computer instructions stored thereon, which, when executed by a processor, implement the above-mentioned short-term power prediction method for new energy stations taking into account power prediction deviations.

[0046] The above technical solution of the embodiment of the present invention has the following beneficial technical effects:

[0047] 1. By considering seasonal factors and daily time factors for calibration, it can better adapt to the changing characteristics of power output of new energy stations under different time conditions. The calibrated prediction value is more in line with the actual power generation situation, effectively reducing the prediction deviation caused by time-related factors and improving the overall prediction accuracy;

[0048] 2. The meteorological conditions in different seasons vary greatly. The conditions such as light, temperature, wind speed, etc. also change at different times of the day. The impact of complex environmental factors on the power of new energy stations is fully considered, so that the prediction model can accurately output results under various working conditions and enhance the adaptability to complex environments;

[0049] 3. Accurate power forecasting helps the grid dispatching department to rationally arrange power generation plans, adjust reserve capacity, better coordinate the power generation ratio of new energy and traditional energy, and ensure the frequency and voltage stability of the grid. BRIEF DESCRIPTION OF THE DRAWINGS

[0050] Figure 1 It is a flow chart of a method for short-term power prediction of a new energy station taking into account power prediction deviation provided by an embodiment of the present invention;

[0051] Figure 2 It is a module block diagram of a short-term power prediction device for a new energy station taking into account power prediction deviation provided by an embodiment of the present invention.

[0052] Reference numerals:

[0053] 1. Meteorological data acquisition module, 2. Power generation prediction module, 3. Power generation calibration module. DETAILED DESCRIPTION

[0054] In order to make the purpose, technical scheme and advantages of the present invention clearer, the present invention is further described in detail below in conjunction with specific embodiments and with reference to the accompanying drawings. It should be understood that these descriptions are only exemplary and are not intended to limit the scope of the present invention. In addition, in the following description, the description of well-known structures and technologies is omitted to avoid unnecessary confusion of the concept of the present invention.

[0055] Please refer to Figure 1 A first aspect of an embodiment of the present invention provides a method for short-term power prediction of a new energy station considering power prediction deviation, comprising the following steps:

[0056] Step S100, obtaining real-time meteorological data of new energy stations.

[0057] Step S200, based on the short-term power prediction model of the new energy station, combined with real-time meteorological data, calculate the short-term power prediction value of the new energy station in the subsequent several detection cycles.

[0058] Step S300: calibrate the short-term power generation prediction value based on the power prediction calibration model of the new energy station to obtain the short-term power generation prediction calibration value of the new energy station.

[0059] Among them, the calibration conditions of the power prediction calibration model include: seasonal factors and daily time factors.

[0060] The above-mentioned short-term power prediction method for new energy stations can cope with changes in meteorological conditions and power consumption patterns by setting calibration conditions in multiple dimensions and taking into account seasonal and daily time factors, thereby improving the accuracy of short-term power predictions and reducing prediction deviations caused by these factors; it also enhances the adaptability of short-term power predictions to meet the characteristics of new energy station power output changes due to different times and environments, making the predicted values ​​more in line with reality.

[0061] Furthermore, the short-term power prediction model based on the new energy station in step S200 is combined with real-time meteorological data to calculate the short-term power prediction value of the new energy station in the subsequent several detection cycles, including:

[0062] Step S210, obtaining historical meteorological data and corresponding historical power generation data of the new energy station.

[0063] Step S220: establishing a correspondence between historical meteorological data and historical power generation data through a short-term power prediction model.

[0064] Step S230, based on the corresponding relationship and combined with the real-time meteorological data, calculate the short-term power generation forecast value of the new energy station.

[0065] By establishing a corresponding relationship using historical meteorological and power generation data, we can discover the patterns between the data required for short-term power prediction of new energy stations, making the short-term power prediction model more in line with the actual power generation characteristics of the station; on this basis, combined with real-time meteorological data to calculate the predicted value, we can fully refer to past experience and patterns, effectively improve the accuracy of power prediction for subsequent detection periods, and provide more reliable data support for subsequent calibration and related work such as grid dispatching and station operation, thereby enhancing the stability and economy of the entire new energy power generation system.

[0066] Specifically, new energy sites include wind turbines and photovoltaic generators, and the corresponding historical meteorological data include wind speed, wind direction, temperature and air pressure of wind turbines and solar irradiance and temperature of photovoltaic generators.

[0067] In a specific implementation of the embodiment of the present invention, the calculation formula of the short-term power generation prediction value of the wind turbine in the short-term power prediction model is:

[0068]

[0069] Among them, P w-p is the short-term power generation prediction value of the wind turbine, ρ is the air density value, A is the swept area of ​​the wind turbine blade, C P is the power coefficient, which is related to the tip speed ratio λ, pitch angle β, wind speed v, and actual temperature T w , actual air pressure p, wind direction is related to the angle θ of the fan main shaft, Δv is the wind speed loss, ΔP turbulence is the power loss caused by turbulence, f yaw (θ) is the wind direction correction function, f control (v,ω) is the fan control strategy function, which is related to the actual wind speed v and the impeller speed ω, f health is the health status factor of the wind turbine, θ is the angle between the actual wind direction and the main axis direction of the wind turbine, v ci is the wind turbine cut-in wind speed, v r is the rated wind speed of the wind turbine, P wr is the rated power of the wind turbine.

[0070] In the above calculation formula, when the wind speed v is lower than the wind turbine cut-in wind speed v ci When the wind speed v is higher than the rated wind speed v of the wind turbine, the wind turbine does not generate electricity. r When the fan output power is the rated power P wr . Power coefficient C P is the tip speed ratio λ, pitch angle β, wind speed v, actual temperature Tw , actual air pressure p, wind direction and the angle θ between the fan shaft, power coefficient C P The specific calculation of can be a complex function obtained by fitting experimental data, which comprehensively considers the influence of multiple factors on the aerodynamic performance of the fan. turbulence The influence of turbulence on power. Turbulence intensity is related to factors such as wind speed, topography, and atmospheric stability. The turbulence intensity distribution can be obtained through field measured data or CFD (computational fluid dynamics) simulation, and then the power loss caused by turbulence can be calculated. Generally speaking, turbulence will cause uneven airflow distribution on the wind turbine blades, reducing the efficiency of wind energy capture. It indicates the proportion of wind speed loss due to factors such as wake effect. The wake effect can be calculated by more complex models, such as the improved Jensen model or CFD simulation results that consider factors such as the distance between wind turbines, wind turbine size, wind direction, etc. Wind direction correction function f yaw (θ) describes the effect of the angle θ between the wind direction and the main axis of the wind turbine on the power output. When the wind turbine is facing the wind direction (θ = 0), f yaw (θ)=1, the power output is maximum; when the wind direction deviates, the power output is based on f yaw The value of (θ) decreases accordingly. The wind turbine adjusts parameters such as pitch angle and generator torque according to wind speed and rotor speed to optimize power output and protect wind turbine equipment. At low wind speeds, the wind energy capture efficiency is maximized by controlling the pitch angle; at high wind speeds, the power output is limited by adjusting the pitch angle and generator torque to prevent wind turbine overload; the wind turbine control strategy function f control (v,ω) can be determined based on the wind turbine control algorithm and actual operating data. Wind turbine health factor f health The impact of the health status of each component of the wind turbine (such as blades, gearbox, generator, etc.) on the power output is comprehensively considered. The health status of the components can be monitored by various sensors installed on the wind turbine (such as vibration sensors, temperature sensors, etc.).

[0071] In the short-term power forecasting process of new energy stations, there are many factors that have a great impact on the power forecast value, mainly changes in meteorological conditions and equipment operating status. For example, compared with the power generation prediction of wind turbines, wind direction and wind speed have a greater impact; compared with photovoltaic generators, solar irradiance, temperature and cloud thickness have a more obvious impact.

[0072] Specifically, when there are sudden changes in wind speed and direction, the power forecast deviation may increase. For example, when the wind direction changes suddenly, the wind turbine needs a certain amount of time to adjust its direction (through the yaw system), and during this adjustment process, the power output will be affected. If the prediction model does not take into account the dynamic process of this wind direction change in a timely manner, deviations will occur. When the wind speed fluctuates, especially when the wind speed is close to the cut-in wind speed or rated wind speed of the wind turbine, the power output is very sensitive to changes in wind speed. If the real-time wind speed data shows that the wind speed fluctuates around the critical value, it is necessary to consider making corresponding fine-tuning to the predicted power. For example, when the wind speed suddenly rises from slightly below the rated wind speed to slightly above the rated wind speed, the predicted power should be adjusted from the proportional increase state to the rated power state.

[0073] In addition, blade wear and icing of the wind turbine will affect the performance of the wind turbine, which will lead to power prediction deviation. The status of the blades can be monitored by sensors installed on the wind turbine. For example, when the sensor detects that the blades have a certain degree of wear, the power prediction value is adjusted according to the degree of wear. Worn blades will reduce the wind energy capture efficiency of the wind turbine, so the predicted power should be appropriately reduced. Mechanical or electrical failures of the wind turbine will also affect the power output. When a fault alarm is detected in the wind turbine, such as excessive generator temperature, gearbox failure, etc., the predicted power of the wind turbine needs to be adjusted to a lower value or directly set to zero (if the fault causes the wind turbine to shut down), and the impact of the wind turbine failure on the power output of the entire wind farm is considered, and the total power prediction value is adjusted accordingly.

[0074] In another specific implementation of the embodiment of the present invention, the calculation formula of the short-term power generation prediction value of the photovoltaic power generation group in the short-term power prediction model is:

[0075]

[0076] Among them, P l-p is the short-term power generation prediction value of the photovoltaic generator set, η(T l ,I,α,β,γ) is the photoelectric conversion efficiency, α is the installation angle of the solar panel, β is the azimuth angle, γ is the solar altitude angle, T l is the actual temperature of the photovoltaic panels of the photovoltaic generator set, I is the actual solar irradiance value, S is the total area of ​​the photovoltaic panels of the photovoltaic generator set, T ref is the preset reference temperature, T co is the temperature coefficient, f shadow (x,y) is the shadow occlusion function, x and y are the distance and height of the occluder respectively, f dirt (d) is the dust coverage function, d is the dust thickness on the panel surface, f aging (t) is the aging function of the solar panel, and t is the service life of the solar panel.

[0077] Specifically, η(T l , I, α, β, γ) is the photoelectric conversion efficiency, which is the panel installation inclination angle α, azimuth angle β, solar altitude angle γ, and the actual temperature value T of the photovoltaic panel of the photovoltaic generator set l and are functions of the actual solar irradiance value I. The relationship between the photoelectric conversion efficiency and the above factors can be obtained by fitting the experimental data. For example, an increase in temperature will lead to a decrease in conversion efficiency, while when the solar irradiance increases within a certain range, the conversion efficiency will have a specific change pattern. At the same time, the installation angle (inclination and azimuth) of the solar panel and the solar altitude angle will affect the effective area and angle of the solar panel receiving the solar irradiance, thereby affecting the conversion efficiency. Preset reference temperature T ref The temperature under standard test conditions is usually taken, such as 25°C.

[0078] Temperature coefficient T co It is a parameter used to measure the degree of influence of temperature change on power of photovoltaic panels. In a physical sense, it describes the temperature change corresponding to the proportion of power reduction when the panel temperature rises by 1°C. The temperature coefficient reflects the temperature characteristics of photovoltaic panel materials and the sensitivity of the entire photovoltaic power generation system to temperature changes. For a given temperature coefficient, it indicates how many degrees of temperature change will cause the power to drop in a certain proportion within a certain temperature range. The size and positive and negative of the temperature coefficient depend on factors such as the material, manufacturing process and system design of the photovoltaic panel. In actual situations, the increase in temperature usually leads to a decrease in the photoelectric conversion efficiency of the panel, thereby reducing the power output, so the temperature coefficient is generally a positive number.

[0079] For example, when the temperature coefficient is 80°C, the actual temperature changes from 25°C (T ref ) to 35℃(T l ), ΔT = 35°C - 25°C = 10°C. At this time, Right now This means that due to a 10°C increase in temperature, the power drops to 0.875 times the original value (at this time, only the temperature factor is considered). Then, this temperature-corrected coefficient is multiplied by the photoelectric conversion efficiency, the total area of ​​the solar panels, and the solar irradiance to obtain the output power of the photovoltaic generator set after considering the temperature effect. In this way, the temperature coefficient plays a key role in correcting the effect of temperature on power in power calculation, so that the power calculation results more accurately reflect the actual situation.

[0080] In addition, for the shadow occlusion function f shadow (x, y), when there is a building blocking the view, the area ratio of the solar panel covered by the shadow is calculated according to the building's position and the sun's azimuth and altitude angle, thereby determining the power reduction ratio; if there is no shadow, f shadow(x, y) = 1, and its value is less than 1 when there is a shadow. Dust will reduce the absorption of solar irradiance by the solar panel. The relationship between the dust coverage and the power reduction ratio is determined through experiments or field monitoring data. When the dust thickness increases, the dust coverage function f dirt (d) gradually decreases, indicating that the power is gradually reduced. Solar panel aging function f aging (t) represents the service life or operating time of the solar panel. As time goes by, the photovoltaic conversion efficiency of the solar panel will decrease and the material performance will decline. The f is determined by studying the aging curve of the solar panel. aging (t), for example, as t increases, f aging (t) gradually decreases, indicating that the power decreases due to aging.

[0081] Compared with photovoltaic power generation units, in the short-term power forecasting process of new energy stations, the factors that have a greater impact on the power forecast value are still changes in meteorological conditions and equipment operating status; for example: solar irradiance, temperature and cloud thickness all have a more obvious impact.

[0082] Specifically, the rapid movement of clouds can cause a sharp change in solar irradiance, which is an important reason for the deviation in power prediction of photovoltaic power generation devices. The rate of change of irradiance can be monitored in real time by high-precision irradiators installed at the site. If the rate of change of irradiance exceeds a certain threshold, it means that there may be cloud cover or sudden weather changes. At this time, the power prediction value is adjusted according to the trend of irradiance change (whether it is a sudden drop or rise) and the type of cloud (thin clouds or thick clouds). For example, for thin cloud cover, the irradiance decreases less, and the power prediction value can be adjusted accordingly according to the actual decrease ratio of irradiance; for thick cloud cover, it may be necessary to consider a larger power adjustment, or even reduce the power prediction value to a lower level until the clouds move away.

[0083] In addition, the operating status of the equipment also has a significant impact on short-term power generation. For example, the surface cleanliness of photovoltaic panels will affect their efficiency in receiving solar irradiance. If there are obstructions such as dust and dirt on the surface, the photoelectric conversion efficiency will be reduced. The cleanliness of the panels can be evaluated by dust sensors installed near the panels or regular cleaning records. The power prediction value is adjusted according to the degree of cleanliness. For example, when there is a lot of dust, the power prediction value is reduced according to the proportion of photoelectric conversion efficiency reduction corresponding to the dust coverage area and thickness. The aging of photovoltaic panels is also an important factor. With the increase of usage time, the photoelectric conversion efficiency of the panels will gradually decrease. By recording the service life and performance attenuation curve of the panels and combining with real-time meteorological data, the power prediction value is calibrated for the aging factor. For example, for panels with a long service life, the predicted power is appropriately reduced to reflect the actual decline in their power generation capacity.

[0084] Furthermore, the power prediction calibration model based on the new energy station in step S300 calibrates the short-term power generation prediction value to obtain the short-term power generation prediction calibration value of the new energy station, including:

[0085] Step S310, obtaining historical power prediction data and corresponding historical power real data of new energy stations.

[0086] First, obtain the historical power forecast data and corresponding historical power real data of the past few years from the database of the new energy site (wind farm or photovoltaic power station), and collect the timestamps (including date and specific time) and meteorological data (such as wind speed, wind direction, solar irradiance, temperature, air pressure, etc.) corresponding to these data. Then, clean the collected data and remove obviously abnormal data points (such as power forecast values ​​or real values ​​beyond the reasonable range, or meteorological data that does not conform to physical laws) to ensure that the forecast data and real data correspond one-to-one in time. If there are missing values, appropriate interpolation methods (such as linear interpolation) can be used to fill them.

[0087] Step S320, based on the power prediction calibration model, calculate the average value of the difference between the historical power prediction data and the corresponding historical power actual data.

[0088] Step S330, when the difference average value is positive, the difference between the short-term power generation prediction value and the difference average value is calculated to obtain the short-term power generation prediction calibration value.

[0089] When the average difference is positive, it means that the historical forecast values ​​are generally too high. Therefore, by subtracting this average difference value, the current short-term power generation forecast value can be adjusted downward to make it closer to the actual possible power output, thereby achieving the purpose of calibration.

[0090] Step S340: when the difference average value is negative, the sum of the short-term power generation prediction value and the absolute value of the difference average value is calculated to obtain the short-term power generation prediction calibration value.

[0091] When the average difference is negative, it means that the historical forecast values ​​are generally low, so it is necessary to adjust the current short-term power generation forecast value upward by adding the absolute value of the average difference value to correct the historical underestimation deviation and make the calibrated forecast value more consistent with the actual power output.

[0092] Furthermore, after obtaining the short-term power generation prediction calibration value in step S200, the method further includes:

[0093] Step S351, calculate the difference average of the corresponding dimension according to the seasonal dimension and the daily time dimension, obtain the change law of the difference average of the corresponding dimension and the overall difference average, and obtain the adjustment coefficient of the difference average of the corresponding dimension and the overall difference average.

[0094] Specifically, the above-mentioned seasonal dimension can divide a year into four seasons: spring, summer, autumn, and winter, according to the local seasonal characteristics. The boundaries of each season can be determined according to a specific date range (for example, spring in the northern hemisphere is from March to May) or meteorological conditions (such as average temperature, sunshine duration, etc.). For each season, all historical power forecast data and real data pairs in the season are screened out. The difference sequence for each season is calculated, and then the spring difference average is calculated. The difference averages for summer, autumn, and winter are calculated in the same way. Take wind turbines as an example: if the spring deviation is positive and the variance is large, it may be caused by frequent wind speed fluctuations in spring and excessive response of the prediction model to wind speed. If the winter deviation is negative and the variance is small, it may be that the low winter temperature affects the performance of the wind turbine. The corresponding adjustment coefficient can be set for the above situation.

[0095] Specifically, the above-mentioned daily time period dimension can divide a day into different time periods, such as night (0-6 o'clock), morning (6-9 o'clock), morning (9-12 o'clock), noon (12-14 o'clock), afternoon (14-18 o'clock), evening (18-21 o'clock) and night (21-24 o'clock). Taking photovoltaic power generation devices as an example, if the deviation in the evening period is negative and the variance is large, it may be that the solar irradiance decreases rapidly in the evening; if the deviation in the noon period is positive and the variance is small, it may be that when the solar irradiance is strongest at noon, the model does not take into account the decrease in photoelectric conversion efficiency caused by the increase in panel temperature. The corresponding adjustment coefficient can be set for the above situation.

[0096] Step S352, adjusting the short-term power prediction value according to the adjustment coefficient to improve the accuracy of the short-term power prediction value.

[0097] Furthermore, the short-term power prediction method of the new energy station considering the power prediction deviation also includes:

[0098] Step S353, respectively obtain the adjustment coefficients of the season dimension and the day time dimension.

[0099] Step S354: calibrate the short-term power prediction values ​​of the wind turbine generator set and the photovoltaic generator set based on the adjustment coefficient.

[0100] Meteorological conditions in different seasons (such as wind speed, solar irradiance, temperature, etc.) have a significant impact on the power output of new energy sites. By calculating the average difference by season, we can clearly understand the characteristics of the power forecast deviation in each season. For example, for wind farms, the power forecast deviation may be different from other seasons in winter due to high wind speeds and frequent changes in wind direction; for photovoltaic power stations, the solar irradiance is strong in summer, but high temperatures may affect the efficiency of the solar panels, causing specific seasonal deviations in power forecasts. Based on the adjustment coefficient of the seasonal dimension, the prediction model can be optimized. If the average difference of a season is positive, it means that the prediction model overestimates the power output as a whole in that season. It may be necessary to adjust the adjustment coefficient in the calibration model or consider more factors related to the season to reduce the prediction value.

[0101] The meteorological conditions and power demand at different times of the day change dynamically. For example, for photovoltaic power stations, the solar irradiance is highest at noon, and the power output usually reaches its peak value. For wind farms, the wind speed distribution may be different during the day and at night, which in turn affects the power output. By calculating the average value of the difference by daily time period, we can understand the power forecast deviation of each time period in detail. For example, in the morning, the power forecast deviation may show a specific change trend due to the gradual increase in solar irradiance or the gradual increase in wind speed. Based on the adjustment coefficient of the daily time period dimension, the calibration model adjustment coefficient of different daily time periods can be adjusted in a targeted manner, which can effectively reduce the power forecast deviation of each time period in the day, improve the accuracy of short-term prediction, and better meet the needs of real-time dispatch of the power system.

[0102] Furthermore, the calculation formula of the short-term power generation prediction calibration value of the wind turbine in the power prediction calibration model is:

[0103]

[0104] Among them, K season-i is the seasonal dimension adjustment coefficient, i is the seasonal number, K daytime-k is the daily time period dimension adjustment coefficient, and k is the corresponding serial number of the daily time period.

[0105] K season-i is the seasonal dimension adjustment coefficient, which is a variable related to the season and can be determined based on statistical analysis of historical data from different seasons. For example, in winter, due to the lower temperature, higher air density, and possibly different wind speed distribution, K season-i The value of may be different in other seasons. Assume that K season-1 =1.0 as the benchmark. After data analysis, it is found that the overall wind turbine power in winter is higher than that in spring (only an example, the actual situation varies from station to station and region). So K in winter season-4=1.1 can comprehensively consider the impact of seasons on factors such as wind speed, air density, solar radiation (indirectly affecting temperature, etc.), and make seasonal adjustments to power forecasts.

[0106] K daytime-k It is the daily time period dimension adjustment coefficient, which reflects the impact of different time periods on the wind turbine power output. For example, during the day, solar radiation may cause the ground temperature to rise, which in turn affects air flow and wind speed. At the same time, electricity demand will also change with people's activities. These factors will affect the wind turbine power output. Taking the noon period as an example, since solar radiation is the strongest, uneven heating of the ground may cause airflow changes, making the wind speed and wind direction have certain special characteristics during this period. Through historical data analysis, if it is found that the wind turbine power during the noon period is 5% higher than other periods on average (also an example), then the K of the noon period daytime-4 =1.05.

[0107] Furthermore, the calculation formula of the short-term power generation prediction calibration value of the photovoltaic generator set in the short-term power prediction model is:

[0108]

[0109] Among them, K season-i is the seasonal dimension adjustment coefficient, i is the seasonal number, K daytime-k is the daily time period dimension adjustment coefficient, and k is the corresponding serial number of the daily time period.

[0110] The adjustment method of each dimension adjustment coefficient of the photovoltaic power generation unit is similar to that of the wind turbine unit, which will not be repeated here.

[0111] Accordingly, please refer to Figure 2 A second aspect of an embodiment of the present invention provides a short-term power prediction device for a new energy station taking into account power prediction deviation, comprising:

[0112] Meteorological data acquisition module 1, which is used to obtain real-time meteorological data of new energy stations;

[0113] The power generation prediction module 2 is used to calculate the short-term power generation prediction value of the new energy station in the subsequent several detection cycles based on the short-term power prediction model of the new energy station and the real-time meteorological data;

[0114] A power generation calibration module 3, which is used to calibrate the short-term power generation prediction value based on the power prediction calibration model of the new energy station to obtain the short-term power generation prediction calibration value of the new energy station;

[0115] The calibration conditions of the power prediction calibration model include: seasonal factors and daily time factors.

[0116] Accordingly, a third aspect of an embodiment of the present invention provides an electronic device, comprising: at least one processor; and a memory connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor executes the above-mentioned short-term power prediction method for new energy stations taking into account power prediction deviations.

[0117] Accordingly, a fourth aspect of an embodiment of the present invention provides a computer-readable storage medium having computer instructions stored thereon, which, when executed by a processor, implement the above-mentioned short-term power prediction method for new energy stations taking into account power prediction deviations.

[0118] The embodiment of the present invention aims to protect a short-term power prediction method for a new energy station taking into account power prediction deviation, comprising the following steps: obtaining real-time meteorological data of the new energy station; calculating the short-term power generation prediction value of the new energy station in several subsequent detection cycles based on the short-term power prediction model of the new energy station in combination with the real-time meteorological data; calibrating the short-term power generation prediction value based on the power prediction calibration model of the new energy station to obtain the short-term power generation prediction calibration value of the new energy station; wherein the calibration conditions of the power prediction calibration model include: seasonal factors and daily time period factors. The above technical scheme has the following effects:

[0119] 1. By considering seasonal factors, daily time factors and working day factors for calibration, it can better adapt to the changing characteristics of power output of new energy sites under different time conditions. The calibrated prediction value is more in line with the actual power generation situation, effectively reducing the prediction deviation caused by time-related factors and improving the overall prediction accuracy;

[0120] 2. The meteorological conditions in different seasons vary greatly, and the conditions such as light, temperature, wind speed, etc. at different times of the day are also changing. The impact of complex environmental factors on the power of new energy stations is fully considered, so that the prediction model can accurately output results under various working conditions and enhance the adaptability to complex environments;

[0121] 3. Accurate power forecasting helps the grid dispatching department to rationally arrange power generation plans, adjust reserve capacity, better coordinate the power generation ratio of new energy and traditional energy, and ensure the frequency and voltage stability of the grid.

[0122] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems, or computer program products. Therefore, the present application may adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment in combination with software and hardware. Moreover, the present application may adopt the form of a computer program product implemented in one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) that include computer-usable program code.

[0123] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0124] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0125] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0126] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the above embodiments, ordinary technicians in the relevant field should understand that the specific implementation methods of the present invention can still be modified or replaced by equivalents. Any modification or equivalent replacement that does 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 short-term power prediction method for a new energy station considering power prediction deviation, characterized in that: The steps include: Obtain real-time meteorological data from new energy stations; Based on the short-term power prediction model of the new energy station, the short-term power prediction value of the new energy station in the subsequent several detection cycles is calculated in combination with the real-time meteorological data; Calibrate the short-term power generation prediction value based on the power prediction calibration model of the new energy station to obtain the short-term power generation prediction calibration value of the new energy station; The calibration conditions of the power prediction calibration model include: seasonal factors and daily time factors.

2. The short-term power prediction method for new energy stations considering power prediction deviation according to claim 1 is characterized in that: The short-term power prediction model based on the new energy station is used to calculate the short-term power prediction value of the new energy station in the subsequent several detection cycles in combination with the real-time meteorological data, including: Obtain historical meteorological data and corresponding historical power generation data of new energy stations; By means of the short-term power prediction model, a corresponding relationship between the historical meteorological data and the historical power generation data is established; Based on the corresponding relationship and in combination with the real-time meteorological data, the short-term power generation forecast value of the new energy station is calculated.

3. The short-term power prediction method for new energy stations considering power prediction deviation according to claim 2 is characterized in that: The new energy station includes: wind turbines and photovoltaic generators; The historical meteorological data include: the wind speed, wind direction, temperature and air pressure corresponding to the wind turbine generator set and the solar irradiance and temperature corresponding to the photovoltaic generator set.

4. The short-term power prediction method for new energy stations considering power prediction deviation according to claim 3 is characterized in that: The calculation formula of the short-term power prediction value of the wind turbine in the short-term power prediction model is: Among them, P w-p is the short-term power generation prediction value of the wind turbine, ρ is the air density value, A is the swept area of ​​the wind turbine blade, C P is the power coefficient, which is related to the tip speed ratio λ, pitch angle β, wind speed v, and actual temperature T w , actual air pressure p, wind direction is related to the angle θ of the fan main shaft, Δv is the wind speed loss, ΔP turbulence is the power loss caused by turbulence, f yaw (θ) is the wind direction correction function, f control (v,ω) is the fan control strategy function, which is related to the actual wind speed v and the impeller speed ω, f health is the health status factor of the wind turbine, θ is the angle between the actual wind direction and the main axis direction of the wind turbine, v ci is the wind turbine cut-in wind speed, v r is the rated wind speed of the wind turbine, P wr is the rated power of the wind turbine.

5. The short-term power prediction method for new energy stations considering power prediction deviation according to claim 3 is characterized in that: The calculation formula of the short-term power prediction value of the photovoltaic generator set in the short-term power prediction model is: Among them, P l-p is the short-term power generation prediction value of the photovoltaic generator set, η(T l ,I,α,β,γ) is the photoelectric conversion efficiency, α is the installation inclination of the solar panel, β is the azimuth, γ is the solar altitude angle, T l is the actual temperature of the photovoltaic panels of the photovoltaic generator set, I is the actual solar irradiance value, S is the total area of ​​the photovoltaic panels of the photovoltaic generator set, T ref is the preset reference temperature, T co is the temperature coefficient, f shadow (x,y) is the shadow occlusion function, x and y are the distance and height of the occluder respectively, f dirt (d) is the dust coverage function, d is the dust thickness on the panel surface, f aging (t) is the aging function of the solar panel, and t is the service life of the solar panel.

6. The short-term power prediction method for new energy stations considering power prediction deviation according to claim 4 or 5 is characterized in that: The short-term power generation prediction value is calibrated based on the power prediction calibration model of the new energy station to obtain the short-term power generation prediction calibration value of the new energy station, including: Obtaining historical power forecast data and corresponding historical power real data of the new energy station; Based on the power prediction calibration model, calculating the average value of the difference between the historical power prediction data and the corresponding historical power real data; When the difference average value is positive, the difference between the short-term power generation prediction value and the difference average value is calculated to obtain the short-term power generation prediction calibration value; When the difference average value is negative, the sum of the short-term power generation prediction value and the absolute value of the difference average value is calculated to obtain the short-term power generation prediction calibration value.

7. The short-term power prediction method for new energy stations considering power prediction deviation according to claim 6 is characterized in that: After obtaining the short-term power generation prediction calibration value, the method further includes: Calculate the difference average of the corresponding dimension according to the seasonal dimension and the daily time period dimension, obtain the change rule of the difference average of the corresponding dimension and the overall difference average, and obtain the adjustment coefficient of the difference average of the corresponding dimension and the overall difference average; The short-term power prediction value is adjusted according to the adjustment coefficient to improve the accuracy of the short-term power prediction value.

8. The short-term power prediction method for new energy stations considering power prediction deviation according to claim 7 is characterized in that: Also includes: Respectively obtain the adjustment coefficients of the season dimension and the day time dimension; The short-term power prediction values ​​of the wind turbine generator set and the photovoltaic generator set are calibrated based on the adjustment coefficient.

9. The short-term power prediction method for new energy stations considering power prediction deviation according to claim 8 is characterized in that: The calculation formula of the short-term power generation prediction calibration value of the wind turbine in the power prediction calibration model is: Among them, K season-i is the seasonal dimension adjustment coefficient, i is the seasonal number, K daytime-k is the daily time period dimension adjustment coefficient, and k is the corresponding serial number of the daily time period.

10. The short-term power prediction method for new energy stations considering power prediction deviation according to claim 8 is characterized in that: The calculation formula of the short-term power prediction calibration value of the photovoltaic generator set in the short-term power prediction model is: Among them, K season-i is the seasonal dimension adjustment coefficient, i is the seasonal number, K daytime-k is the daily time period dimension adjustment coefficient, and k is the corresponding serial number of the daily time period.

Images