Solar power generation prediction method, system, equipment and medium for aircraft

Abstract

The invention provides a solar power generation prediction method, system and device for an aircraft and a medium, and relates to the technical field of near space aircrafts, and the method comprises the steps: obtaining target operation environment state data corresponding to a target aircraft, the target operation environment state data at least comprises sun-earth distance data, earth orbit kinetic parameters, aircraft shielding parameters, environment temperature parameters and aircraft sensor data; the aircraft shielding parameter represents the shielding range of the solar cell by the component of the target aircraft; the target aircraft is a near space aircraft; and inputting the target operation environment state data into a solar power physical model and a solar power prediction model in the solar power generation prediction comprehensive model to obtain solar cell array output prediction power corresponding to the target aircraft output by the solar power generation prediction comprehensive model. According to the invention, more accurate output prediction power of the solar cell array can be obtained.

Description

Technical Field

[0001] This invention relates to the field of near-space vehicle technology, and in particular to a method, system, device and medium for predicting solar power generation for aircraft. Background Technology

[0002] One of the most common methods for forecasting solar power generation is to use meteorological data, particularly solar radiation data, to estimate the output power of solar panels. Traditional meteorological models rely on weather forecast data, such as temperature, humidity, wind speed, and cloud cover, to predict the intensity of solar radiation over a future period. These methods have a certain degree of accuracy, especially under relatively stable weather conditions, but their predictive performance is poor for sudden changes in weather conditions (such as rapid cloud movement or sudden thunderstorms).

[0003] Near-space long-endurance low-speed aircraft, including solar-powered aircraft, stratospheric airships, and high-altitude balloons, offer several unique advantages, such as providing low-cost continuous observation, acting as relays in communications, and potential applications in emergency rescue operations. To achieve extended flight times, near-space long-endurance low-speed aircraft rely on renewable energy systems, with solar energy considered the most ideal energy source.

[0004] However, current predictions for solar power generation by near-space low-speed aircraft have low accuracy. One key reason is that existing physical models cannot accurately predict the actual output power of solar panels at midday. For the energy design and mission planning of near-space low-speed aircraft, inaccurate predictions could lead to mission failure in severe cases. Therefore, there is an urgent need for a method, system, equipment, and medium for predicting solar power generation for aircraft to solve these problems. Summary of the Invention

[0005] To address the problems existing in the prior art, the present invention provides a method, system, device and medium for predicting solar power generation for aircraft.

[0006] This invention provides a method for predicting solar power generation for aircraft, comprising: Acquire target operating environment status data corresponding to the target spacecraft, wherein the target operating environment status data includes at least Earth-Sun distance data, Earth orbital dynamics parameters, spacecraft shading parameters, ambient temperature parameters, and spacecraft sensor data; the spacecraft shading parameters represent the shading range of the solar cells by the components of the target spacecraft; the target spacecraft is a near-space spacecraft; The target operating environment status data is input into the solar power physical model and the solar power prediction model in the solar power prediction integrated model, respectively, to obtain the predicted output power of the solar cell array corresponding to the target aircraft output by the solar power prediction integrated model; The solar power physics model is used to obtain the output power of the first solar cell based on total solar radiation, effective shading area of ​​the solar cell, and temperature effect. The total solar radiation, effective shading area of ​​the solar cell, and temperature effect are calculated based on the target operating environment state data. The solar power prediction model is trained based on a deep learning model and is used to predict the output power of the second solar cell based on the input target operating environment state data. The solar power generation prediction comprehensive model is used to adjust the output power of the first solar cell based on the output power of the second solar cell to obtain the predicted output power of the solar cell array.

[0007] According to the present invention, a method for predicting solar power generation for aircraft is provided, wherein the solar power physical model includes a solar radiation model, a shading model, a temperature effect model, and an output power model, wherein: The solar radiation model is used to calculate the total solar radiation corresponding to the location of the target spacecraft at the current time, based on the Earth-Sun distance data, the Earth orbital dynamics parameters, the spacecraft's spatial position, and the current time. The shadow model is used to calculate the shadow occlusion area of ​​the aircraft balloon and the shadow occlusion area of ​​the aircraft flange according to the aircraft occlusion parameters, and to obtain the effective solar cell occlusion area of ​​the target aircraft at the current time based on the shadow occlusion area of ​​the aircraft balloon and the shadow occlusion area of ​​the aircraft flange. The effective solar cell occlusion area represents the area corresponding to the solar cell area on the aircraft that is not occluded by the aircraft balloon and the aircraft flange. The temperature effect model is used to determine the real-time temperature data and reference temperature data corresponding to the location of the target aircraft at the current time based on the ambient temperature parameters, and to calculate the temperature effect based on the real-time temperature data and the reference temperature data. The output power model is used to calculate the output power of the first solar cell based on the total solar radiation, the effective shading area of ​​the solar cell, and the temperature effect.

[0008] According to the solar power generation prediction method for aircraft provided by the present invention, the calculation formula of the solar radiation model is as follows: ; ; ; ; ; ; in, This represents the total solar radiation. This indicates the intensity of direct solar radiation. Indicates the intensity of reflected radiation. Indicates the solar altitude angle. Indicates sunrise time. Indicates sunset time. Indicates reflectance on sunny days. This represents the corrected extraterrestrial solar radiation. Indicates atmospheric transmittance. Indicates air quality, This indicates the atmospheric pressure at the location of the aircraft. This represents the atmospheric pressure at sea level. Indicates extraterrestrial solar radiation; This represents the eccentricity factor calculated from the Earth-Sun distance data.

[0009] According to the solar power generation prediction method for aircraft provided by the present invention, the calculation formula of the shadow model is as follows: ; ; ; in, This indicates the area of ​​the shadow cast by the aircraft balloon. This represents the area directly exposed to sunlight. Indicates the length of the solar cell array. Indicates the width of the solar cell array. This indicates the area of ​​the aircraft flange that is shaded. Indicates the flange shadow angle. Indicates the first critical angle of shadow. Indicates the second critical angle of shadow. Indicates the length of the aircraft flange. Indicates the width of the aircraft flange. This indicates the effective shading area of ​​the solar cell.

[0010] According to the solar power generation prediction method for aircraft provided by the present invention, the calculation formula of the temperature effect model is as follows: ; in, This refers to the temperature effect. This indicates the temperature effect corresponding to the reference temperature data. Indicates the temperature coefficient. This refers to the real-time temperature data. This refers to the reference temperature data; The calculation formula for the output power model is as follows: ; in, This indicates the output power of the first solar cell. This indicates the projection rate of the outer layer of the solar array.

[0011] According to the present invention, a method for predicting solar power generation for aircraft is provided, wherein the solar power prediction model is trained through the following steps: Acquire the sample operating environment status data and the actual output power of the solar panel cell corresponding to the sample operating environment status data; The sample operating environment status data is input into the solar power physical model to obtain the solar cell sample output power output by the solar power physical model; Based on the error between the actual output power of the solar panel cell and the output power of the solar cell sample, a solar panel cell output power error label corresponding to the sample operating environment status data label is constructed. A training sample set is constructed based on the sample operating environment status data labeled with solar panel cell output power error tags; The deep learning model is trained based on the training sample set to obtain the solar power prediction model. The deep learning model is constructed based on the improved Blur informer model, in which a low-pass filter is set between the max pooling layer and the downsampling layer.

[0012] The present invention also provides a solar power generation prediction system for aircraft, comprising: The data acquisition module is used to acquire target operating environment status data corresponding to the target spacecraft. The target operating environment status data includes at least Earth-Sun distance data, Earth orbit dynamics parameters, spacecraft shading parameters, ambient temperature parameters, and spacecraft sensor data. The spacecraft shading parameters represent the shading range of the solar cells by the target spacecraft's components. The solar power prediction module is used to input the target operating environment status data into the solar power physical model and the solar power prediction model in the solar power prediction integrated model, respectively, to obtain the predicted output power of the solar cell array corresponding to the target aircraft output by the solar power prediction integrated model; The solar power physics model is used to obtain the output power of the first solar cell based on total solar radiation, effective shading area of ​​the solar cell, and temperature effect. The total solar radiation, effective shading area of ​​the solar cell, and temperature effect are calculated based on the target operating environment state data. The solar power prediction model is trained based on a deep learning model and is used to predict the output power of the second solar cell based on the input target operating environment state data. The solar power generation prediction comprehensive model is used to adjust the output power of the first solar cell based on the output power of the second solar cell to obtain the predicted output power of the solar cell array.

[0013] The present invention also provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the solar power generation prediction method for aircraft as described above.

[0014] The present invention also provides a non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the solar power generation prediction method for aircraft as described above.

[0015] The present invention also provides a computer program product, including a computer program that, when executed by a processor, implements the solar power generation prediction method for aircraft as described above.

[0016] The present invention provides a method, system, device, and medium for predicting solar power generation for aircraft. By acquiring the Earth-Sun distance, Earth orbit dynamics parameters, aircraft shading parameters, ambient temperature, and aircraft sensor data of the target near-space aircraft, these data are then input into two sub-models in a comprehensive solar power generation prediction model. The comprehensive model then combines the outputs of the two sub-models and adjusts the first output power of the physical model by using the second output power output by the deep learning model, thereby obtaining a more accurate predicted output power of the solar cell array. Attached Figure Description

[0017] To more clearly illustrate the technical solutions in this invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of this invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0018] Figure 1 A schematic flowchart illustrating the solar power generation prediction method for aircraft provided by the present invention; Figure 2 A schematic diagram of the structure of the solar power generation prediction system for aircraft provided by the present invention; Figure 3 This is a schematic diagram of the structure of the electronic device provided by the present invention. Detailed Implementation

[0019] To make the objectives, technical solutions, and advantages of this invention clearer, the technical solutions of this invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of this invention. All other embodiments obtained by those skilled in the art based on the embodiments of this invention without creative effort are within the scope of protection of this invention.

[0020] In the field of solar power generation forecasting, the traditional approach mainly involves analyzing historical meteorological data and combining it with radiative transfer models to predict short-term trends in solar radiation, thereby estimating the power generation of solar power plants. This method has a certain degree of accuracy when weather changes are relatively stable, but its predictive effectiveness is often poor when faced with sudden changes in weather conditions, such as rapid cloud movement or sudden thunderstorms.

[0021] In recent years, with the rise of artificial intelligence technology, machine learning and deep learning have been widely used in solar power generation forecasting. By training neural network models or performing regression analysis, machine learning methods can extract patterns from large amounts of historical meteorological data and predict solar power output. In particular, deep learning algorithms based on big data, such as Convolutional Neural Networks (CNNs) and Recurrent Neural Networks (RNNs), can effectively handle complex spatiotemporal relationships and significantly improve prediction accuracy. For example, RNNs are good at processing time series data and are suitable for predicting short-term and long-term power generation changes; while CNNs can handle multidimensional data, especially data related to meteorological images, further improving prediction accuracy.

[0022] In addition, there is a class of solar power generation prediction methods based on physical principles. These methods mainly rely on astronomical and meteorological physical models, predicting solar radiation intensity and power generation through detailed calculations of radiative transfer equations, the relative positions of the Earth and the Sun, and the scattering and absorption processes of solar radiation in the atmosphere. However, these methods involve complex calculations, typically requiring significant computational power, and their applicability may vary across different regions and time periods. Therefore, these methods are usually used in conjunction with other methods to improve the reliability and practicality of the predictions.

[0023] For near-space long-endurance, low-speed aircraft, such as solar-powered aircraft, stratospheric airships, and high-altitude balloons, solar energy is the ideal energy source. However, existing physical models are insufficient in predicting the actual output power of solar panels during midday, posing a challenge to the energy design, mission planning, and even safe flight of these aircraft. Accurately predicting the output power of stratospheric balloon solar panels is crucial, affecting not only energy security but also mission planning.

[0024] Current research on solar power generation prediction has considered various influencing factors, such as solar panel temperature, wind speed, illumination angle, shading effect, and solar radiation intensity. However, most studies remain limited to basic models. Furthermore, existing research typically focuses only on standard flight attitudes, neglecting the oscillation phenomena commonly observed in stratospheric balloons during actual flight. Therefore, a more comprehensive consideration of various influencing factors is needed to improve the accuracy of solar power generation prediction and provide strong support for the energy design and mission planning of near-space long-endurance, low-speed vehicles.

[0025] To address the problems existing in the prior art, this invention provides a solar power generation prediction method suitable for near-space vehicles. By combining a physical model and a Blur Informer deep learning model, a comprehensive model for predicting solar power generation from high-altitude balloons is constructed. The physical model optimizes for issues not considered in existing models, particularly addressing the multi-peak problem in Maximum Power Point Tracking (MPPT) and the impact of equipment shadows (such as shadows caused by equipment flanges) on the solar array. Simultaneously, the model is validated using actual flight data, overcoming the lack of real-world data support in existing research. The Blur Informer model in this invention, by introducing an anti-aliasing layer (i.e., a low-pass filter placed between the max-pooling layer and the downsampling layer), improves the model's ability to capture complex nonlinear factors, significantly enhancing overall prediction accuracy.

[0026] Figure 1This is a flowchart illustrating the solar power generation prediction method for aircraft provided by the present invention, as shown below. Figure 1 As shown, the present invention provides a method for predicting solar power generation for aircraft, comprising: Step 101: Obtain target operating environment status data corresponding to the target spacecraft. The target operating environment status data includes at least Earth-Sun distance data, Earth orbit dynamic parameters, spacecraft shading parameters, ambient temperature parameters, and spacecraft sensor data. The spacecraft shading parameters represent the shading range of the solar cells by the components of the target spacecraft. The target spacecraft is a near-space spacecraft.

[0027] In this invention, the target aircraft is a near-space aircraft. These aircraft typically fly at the edge of the Earth's atmosphere or in higher altitudes, and are characterized by long loiter times and high coverage. They are commonly used for tasks such as communication relay, meteorological observation, and scientific experiments. To fully understand and simulate the operating environment of the target aircraft, a series of data needs to be collected. This data is crucial for analyzing the aircraft's performance and predicting its energy needs (such as solar power generation).

[0028] Specifically, the Earth-Sun distance refers to the average distance between the Earth and the Sun. It affects the intensity of solar radiation reaching the Earth (and spacecraft). Since the Earth's orbit around the Sun is elliptical, the Earth-Sun distance varies, which has a significant impact on the prediction of solar power generation.

[0029] Earth's orbital dynamics parameters describe the Earth's motion in the solar system, including its rotational speed, orbital speed, and orbital inclination. These parameters are used to understand how solar radiation varies with time and location, and how spacecraft are positioned relative to the sun.

[0030] Aircraft shading parameters refer to the extent to which components of the target aircraft (such as balloons, wings, fuselages, antennas, etc.) obstruct the solar cells. The degree of shading by these components affects the amount of solar radiation received by the solar cells. These shading parameters include the shape, size, and location of the obstructing objects, as well as their changes with the aircraft's attitude. It should be noted that this invention uses high-altitude balloons or stratospheric airships for illustration; therefore, in addition to considering the obstruction from equipment flanges, the shading from balloons on the solar cells must also be taken into account.

[0031] Ambient temperature affects the efficiency and lifespan of solar cells. As the temperature rises, the conversion efficiency of solar cells may decrease. Therefore, ambient temperature is a factor that needs to be considered when predicting solar power generation.

[0032] In addition, near-space vehicles are equipped with various sensors to monitor the vehicle's status and environmental conditions. These sensor data include the vehicle's attitude, speed, altitude, temperature, and pressure, which are crucial for validating and calibrating predictive models and for adjusting the vehicle's energy management strategies in real time.

[0033] Step 102: Input the target operating environment status data into the solar power physical model and the solar power prediction model in the solar power prediction integrated model, respectively, to obtain the predicted output power of the solar cell array corresponding to the target aircraft output by the solar power prediction integrated model; The solar power physics model is used to obtain the output power of the first solar cell based on total solar radiation, effective shading area of ​​the solar cell, and temperature effect. The total solar radiation, effective shading area of ​​the solar cell, and temperature effect are calculated based on the target operating environment state data. The solar power prediction model is trained based on a deep learning model and is used to predict the output power of the second solar cell based on the input target operating environment state data. The solar power generation prediction comprehensive model is used to adjust the output power of the first solar cell based on the output power of the second solar cell to obtain the predicted output power of the solar cell array.

[0034] In this invention, the target operating environment state data (including Earth-Sun distance data, Earth orbital dynamics parameters, spacecraft shading parameters, ambient temperature parameters, and spacecraft sensor data) obtained in the above embodiments are input into two sub-models of the comprehensive solar power generation prediction model: the solar power physics model and the solar power prediction model. The solar power physics model is mainly used to calculate the output power of the solar cells based on physical principles. This model considers three main factors: total solar radiation, effective shading area of ​​the solar cells, and temperature effect. Specifically, total solar radiation is calculated based on the target operating environment state data (such as Earth-Sun distance and Earth orbital dynamics parameters), representing the intensity of solar radiation reaching the surface of the spacecraft's solar cells; the effective shading area of ​​the solar cells is calculated based on the spacecraft shading parameters, representing the effective area of ​​the solar cells that actually receives solar radiation after being shaded by spacecraft components; and the temperature effect considers the impact of ambient temperature on the efficiency of the solar cells. Using these three factors, the solar power physics model can calculate the output power of the first solar cell.

[0035] The solar power prediction model is trained using a deep learning model. It learns from a large amount of historical data and predicts the output power error of the solar cells, i.e., the output power of the second solar cell. Furthermore, the integrated solar power prediction model combines the outputs of the physical solar power model and the solar power prediction model, adjusting the output power of the first solar cell (from the physical model) based on the output power of the second solar cell (from the deep learning model). Finally, the integrated solar power prediction model outputs the predicted output power of the solar array corresponding to the target aircraft. By combining the accuracy of the physical model and the fitting ability of the deep learning model, the accuracy and reliability of the prediction are improved.

[0036] The solar power generation prediction method for aircraft provided by this invention obtains the Earth-Sun distance, Earth orbit dynamics parameters, aircraft shading parameters, ambient temperature, and aircraft sensor data of the target near-space aircraft. Then, these data are input into two sub-models in the solar power generation prediction integrated model. The integrated model then combines the outputs of the two sub-models and adjusts the first output power of the physical model by using the second output power output by the deep learning model, thereby obtaining a more accurate predicted output power of the solar cell array.

[0037] Based on the above embodiments, the solar power physical model includes a solar radiation model, a shadowing model, a temperature effect model, and an output power model, wherein: The solar radiation model is used to calculate the total solar radiation corresponding to the location of the target spacecraft at the current time, based on the Earth-Sun distance data, the Earth orbital dynamics parameters, the spacecraft's spatial position, and the current time. The shadow model is used to calculate the shadow occlusion area of ​​the aircraft balloon and the shadow occlusion area of ​​the aircraft flange according to the aircraft occlusion parameters, and to obtain the effective solar cell occlusion area of ​​the target aircraft at the current time based on the shadow occlusion area of ​​the aircraft balloon and the shadow occlusion area of ​​the aircraft flange. The effective solar cell occlusion area represents the area corresponding to the solar cell area on the aircraft that is not occluded by the aircraft balloon and the aircraft flange. The temperature effect model is used to determine the real-time temperature data and reference temperature data corresponding to the location of the target aircraft at the current time based on the ambient temperature parameters, and to calculate the temperature effect based on the real-time temperature data and the reference temperature data. The output power model is used to calculate the output power of the first solar cell based on the total solar radiation, the effective shading area of ​​the solar cell, and the temperature effect.

[0038] In this invention, the solar power physics model includes a solar radiation model, a shading model, a temperature effect model, and an output power model to accurately predict the output power of solar cells.

[0039] Specifically, the solar radiation model is used to calculate the total solar radiation corresponding to the target spacecraft's location at the current time. This model relies on Earth-Sun distance data, Earth's orbital dynamics parameters, the spacecraft's spatial position, and the current time. By comprehensively considering factors such as the relative position between the Earth and the Sun, orbital dynamics, and seasonal variations, the solar radiation model can calculate the direct and diffuse radiation intensity at the altitude of the stratospheric spacecraft. It should be noted that since high-altitude balloons fly in the atmosphere above 20 kilometers, the influence of clouds is negligible, but the model will still accurately calculate the impact of atmospheric transmittance on radiation intensity.

[0040] A shadow model is used to calculate the effective area of ​​solar cell shading on a target aircraft at the current time. Its input parameters are primarily aircraft shading parameters, including the shading of the solar cells by the aircraft balloon (if the aircraft is a solar-powered aircraft, balloon shading is not considered) and the aircraft flange. In this invention, the shadow model uses geometric methods to analyze the shading of the solar array by the high-altitude balloon and its auxiliary equipment (such as the flange). The shadow model calculates the shadow area based on different shading angles and evaluates the impact of these shadows on the sunlight received by the solar cells. Finally, the shadow model derives the effective area of ​​solar cell shading, which is the area corresponding to the unshaded solar cell region on the aircraft.

[0041] A temperature effect model is used to calculate the impact of temperature on the output power of solar cells. This model relies on ambient temperature parameters, including real-time temperature data of the target spacecraft's current location and reference temperature data. In this invention, the temperature effect model introduces a temperature coefficient to analyze the changes in the conversion efficiency of the solar array under different temperature conditions. Since temperature fluctuations directly affect the output power of solar cells, especially in high-altitude, low-temperature environments, the temperature effect model ensures accurate predictions even under flight conditions with large temperature differences. Furthermore, the temperature effect model can dynamically adjust temperature parameters to optimize the performance of the solar array in different environments.

[0042] The output power model comprehensively considers the impact of factors such as solar radiation intensity, shading effect, and temperature variation on the output power of solar cells. By incorporating these factors into the calculation of the output power model, it can accurately predict the real-time power output of solar cells and effectively solve the multi-peak problem that may occur during maximum power point tracking (MPPT). This allows for accurate prediction of the output power of solar cells under different environmental conditions, ensuring stable output of solar cells under partial shading or complex lighting conditions.

[0043] In this invention, the process of constructing a solar energy physics model includes four key steps. First, a solar radiation model is established, which considers the changes in the distance between the Earth and the Sun, orbital dynamics, and the influence of different seasons and latitudes on solar radiation intensity, enabling accurate calculation of direct and diffuse radiation. Second, a shading model is established to analyze the shading effect of aircraft balloons (if any) and their auxiliary equipment (such as flanges) on the solar array, assessing their impact on sunlight reception by calculating the shading angle and area. Third, a temperature effect model is established to dynamically adjust the power output prediction to address changes in conversion efficiency caused by temperature fluctuations. Finally, an output power model is established, integrating solar radiation, shading effects, and temperature changes, using the maximum power point tracking (MPPT) algorithm to predict the output power of the solar array under complex operating conditions, and resolving the multi-peak problem under partial shading conditions to ensure array output stability.

[0044] In this invention, the main input energy source for the solar cell array is solar radiation. When calculating the solar radiation intensity on the Earth's surface, it is necessary to consider the distance between the Sun and the Earth, as well as the Earth's orbital dynamics. Furthermore, it is also necessary to consider factors such as the position of the stratospheric vehicle (or balloon), the current time, and the attenuation of radiation in the Earth's atmosphere and clouds, in order to ensure an accurate estimate of the solar radiation intensity.

[0045] In this invention, solar radiation reaching the Earth's surface needs to pass through the atmosphere to reach the stratospheric balloon. Since the altitude of a stratospheric vehicle is typically 20 kilometers, exceeding the height of clouds, the influence of weather phenomena such as clouds and rain need not be considered. Instead, the influence of atmospheric transmittance at different solar zenith angles on radiation intensity must be taken into account.

[0046] Specifically, in this invention, the average Earth-Sun distance is one astronomical unit (AU) = 149,597,890 km. Since the Earth's orbit around the Sun is elliptical, the actual Earth-Sun distance varies throughout the year. Therefore, when calculating the eccentricity correction factor, it needs to be calculated based on the Earth-Sun distance data at the current time. The eccentricity correction factor can be calculated using the following formula. : ; in, This represents the Earth-Sun distance data.

[0047] Furthermore, before calculating the theoretical solar power generation value, it is also necessary to calculate the solar altitude angle. The specific process is as follows: First, determine the solar declination angle. and solar hour angle Declination is the angle between the sun's rays when they are directly overhead at the local meridian and the equatorial plane. North latitude is taken as positive, and its range is [range missing]. declination angle The calculation formula is: ; in, It is 0.006918. It is 0.399912. It is 0.070257. It is 0.006758. It is 0.000907. It is 0.002697. It is 0.001480; The sun angle is represented by the formula for calculating it: ; in, The total number of days in a year. This refers to the number of days in that year. In a normal year, =365, The value range is 1 to 365; in leap years, =366, The value range is from 1 to 366.

[0048] Furthermore, solar altitude angle The calculation formula is: ; in, This represents latitude, with North latitude as positive, and its value ranges from -90° to 90°. It takes into account spatial and temporal location and atmospheric effects, representing extraterrestrial solar radiation. =1367W / m², requires eccentricity factor. After making corrections, the corrected extraterrestrial solar radiation is obtained. .

[0049] Furthermore, since stratospheric vehicles operate at an altitude of 20 km (above the cloud top), they do not need to consider weather phenomena such as clouds and rain, but they do need to calculate the atmospheric transmittance at different zenith angles. : ; ; in, This represents the atmospheric pressure at the location of the aircraft (i.e., an altitude of 20 km). The formula for calculating the intensity of direct solar radiation is: .

[0050] In this invention, although scattered radiation exists, at an altitude of 20 km, only the downward scattering above the solar cell array participates in photoelectric conversion. This downward scattering at this point is less than 1% of the direct radiation and can therefore be ignored. However, reflected radiation provides significant radiative flux; therefore, the intensity of reflected radiation needs to be calculated based on the reflectivity under clear skies. : ; Among them, sunny reflectance The reflectance is approximately 0.18, and for cloudy conditions, it is approximately 0.57. Based on the above embodiments, and using the calculation formulas in the above embodiments, the calculation formula for the solar radiation model is as follows: ; ; ; ; ; ; in, This represents the total solar radiation. This indicates the intensity of direct solar radiation. Indicates the intensity of reflected radiation. Indicates the solar altitude angle. Indicates sunrise time. Indicates sunset time. Indicates reflectance on sunny days. This represents the corrected extraterrestrial solar radiation. Indicates atmospheric transmittance. Indicates air quality, This indicates the atmospheric pressure at the location of the aircraft. This represents the atmospheric pressure at sea level. Indicates extraterrestrial solar radiation; This represents the eccentricity factor calculated from the Earth-Sun distance data. For other In cases where the solar altitude angle is less than 0, the predicted output is 0.

[0051] Furthermore, flanges and other equipment on the aircraft (hereafter referred to as flanges in this embodiment) may cast shadows on the solar cell array. Different angle symbols This indicates the relative position of the flange shadow and the solar cell array; that is, changes in the shadow angle affect the amount of sunlight received by the solar panel. In this invention, for different conditions, the longer the flange shadow, the greater the shading on the solar panel. When the shadow angle is greater than a critical angle, the shadow does not affect the solar cell array; when the shadow angle is equal to the critical angle, the shadow just touches the edge of the solar cell array; when the shadow angle is less than the critical angle, the shadow completely covers the solar panel, affecting its energy output.

[0052] In this invention, both the shadow of the aerostat balloon (if present) and the shadow of the flange are considered simultaneously. The array area directly exposed to sunlight receives all direct light, while the area covered by the aerostat balloon shadow receives less direct light, and the flange shadow completely blocks light. To simplify calculations, the flange shading portion is equivalent to a cube attached to the bottom of the aerostat. Since the aerostat balloon is much larger than the array, the shadow edge can be approximated by a straight line. Because the solar cell array carried by the aerostat balloon is cross-shaped, it is considered a square with side length L for ease of calculation. Simultaneously, assuming the aerostat balloon is circular, the shadow shading areas corresponding to the aerostat balloon shading and flange shading are calculated based on the relationship between the area directly exposed to sunlight and the length of the solar cell array, and the relationship between the flange shadow angle and the critical shadow angle. Thus, the effective area of ​​the currently unshaded solar cells is obtained based on the shadow shading areas corresponding to the aerostat balloon shading and flange shading. It should be noted that for solar-powered aircraft, only the influence of the aircraft flange shadow needs to be considered. Based on the above embodiments, the calculation formula for the shadow model is constructed as follows: ; ; ; in, This indicates the area of ​​the shadow cast by the aircraft balloon. This represents the area directly exposed to sunlight. Indicates the length of the solar cell array. Indicates the width of the solar cell array. This indicates the area of ​​the aircraft flange that is shaded. Indicates the flange shadow angle. Indicates the first critical angle of shadow. Indicates the second critical angle of shadow. Indicates the length of the aircraft flange. Indicates the width of the aircraft flange. This indicates the effective shading area of ​​the solar cell.

[0053] Furthermore, in this invention, the key to establishing the temperature effect model lies in analyzing the impact of temperature changes on the output power of the solar cell array and ensuring accurate evaluation of array performance under different temperature conditions by dynamically adjusting the prediction process. Subsequently, based on real-time temperature data and the temperature effect model, the output power of the solar cell array is dynamically calculated, and the accuracy of the model is verified through experiments under different temperature conditions. The temperature coefficient and parameter settings are optimized to improve the robustness of the model under extreme temperature conditions. Based on the above embodiments, the calculation formula for the temperature effect model is constructed as follows: ; in, This refers to the temperature effect. This indicates the temperature effect corresponding to the reference temperature data. Indicates the temperature coefficient. This refers to the real-time temperature data. This refers to the reference temperature data; The calculation formula for the output power model is as follows: ; in, This indicates the output power of the first solar cell. This indicates the projection rate of the outer layer of the solar array.

[0054] In this invention, when part of the solar cell array on the aircraft is shaded, the influence of the correlation coefficient calculated by the above physical model on the output power of the solar cell array is considered, thereby obtaining a more accurate solar cell output power.

[0055] Based on the above embodiments, the solar power prediction model is obtained through the following steps: Acquire the sample operating environment status data and the actual output power of the solar panel cell corresponding to the sample operating environment status data; The sample operating environment status data is input into the solar power physical model to obtain the solar cell sample output power output by the solar power physical model; Based on the error between the actual output power of the solar panel cell and the output power of the solar cell sample, a solar panel cell output power error label corresponding to the sample operating environment status data label is constructed. A training sample set is constructed based on the sample operating environment status data labeled with solar panel cell output power error tags; The deep learning model is trained based on the training sample set to obtain the solar power prediction model. The deep learning model is constructed based on the improved Blur informer model, in which a low-pass filter is set between the max pooling layer and the downsampling layer.

[0056] In this invention, an improved Blur Informer model is used as the deep learning model. By introducing an anti-aliasing layer into the Blur Informer model, the influence of complex nonlinear factors during high-altitude balloon flight is optimized through deep learning. Existing time series prediction models, such as Informer, while possessing strong time series processing capabilities, are prone to prediction errors under high-frequency perturbations such as attitude changes and shadowing effects. The improved Blur Informer in this invention, by modifying the max pooling layer and introducing a fuzzy pooling layer, effectively reduces signal aliasing and improves the model's ability to capture subtle changes. This mechanism is particularly significant in the context of attitude changes during balloon flight.

[0057] The key innovation of the Blur Informer model in this invention lies in its anti-aliasing mechanism. Existing models' max-pooling layers tend to overlook subtle changes when handling translation invariance, leading to significant output fluctuations from small input variations. To address this issue, the improved Blur Informer model incorporates a low-pass filter between the max-pooling layer and the downsampling process, achieving anti-aliasing. This not only improves the model's stability under complex flight conditions but also enhances its ability to handle multi-peak problems, ensuring the continuity and accuracy of predictions.

[0058] Furthermore, in this invention, the improved Blur Informer model incorporates a multi-head attention mechanism, combining extensive flight data and environmental features (such as solar radiation, temperature, and wind speed) to effectively learn the impact of different environmental factors on the output power of the solar array. In addition, through SHAP (SHapley Additive exPlanations) analysis, the BlurInformer model can explain the contribution of each input feature to the prediction results, providing clear guidance for further model optimization.

[0059] Furthermore, in this invention, the integrated solar power generation prediction model consists of a physical model and a Blur Informer deep learning model, used to solve complex nonlinear problems in high-altitude balloon solar power generation prediction. The physical model focuses on accurately calculating the impact of physical factors such as solar radiation, shading effects, and temperature changes on the output power of the solar array using mathematical formulas. This physical model optimizes the multi-peak problem in maximum power point tracking (MPPT) and the impact of equipment such as flanges on illumination, thereby providing accurate basic prediction data.

[0060] However, relying solely on physical models cannot fully capture complex factors such as attitude changes and environmental fluctuations during flight. Therefore, the comprehensive solar power generation prediction model introduces the Blur Informer deep learning model, which further optimizes the prediction error of the physical model by learning the potential nonlinear relationships in flight data. The Blur Informer in this invention reduces signal aliasing through an anti-aliasing layer, making it particularly suitable for scenarios with drastic attitude changes or variable environmental factors, thus becoming more sensitive to subtle changes and significantly improving prediction accuracy. Furthermore, the Blur Informer possesses a multi-head attention mechanism and time series processing capabilities, enabling it to effectively learn and capture the dynamic changes of multi-dimensional input features such as solar radiation intensity, temperature fluctuations, and wind speed. Compared to traditional time series prediction models, it effectively solves the MPPT multi-peak problem, allowing the solar array to maintain stable output power even under partial shading or complex lighting conditions. This invention applies an improved Blur Informer model, introducing an anti-aliasing layer to enhance the ability to capture complex nonlinear factors such as aircraft attitude changes (e.g., airship attitude changes), refining the prediction error of the physical model. In addition, in this invention, actual flight data is used to train and validate the integrated model. By comparing the prediction results with the actual data, the accuracy of the model is evaluated. Finally, SHAP analysis is used to identify key features that affect the prediction, providing guidance for further optimization of the physical model and improving the overall prediction performance.

[0061] For the training process of deep learning models, this invention requires collecting a series of sample operating environment state data. This data may include various parameters related to the solar cell's operating environment, such as the Earth-Sun distance, Earth orbital dynamics parameters, spacecraft shading parameters, and ambient temperature. Simultaneously, it is necessary to obtain the actual output power of the solar cells corresponding to these sample operating environment state data; that is, the power actually generated by the solar cells under these conditions in the actual operating environment.

[0062] Furthermore, the collected sample operating environment data is input into the solar power physics model. Based on this input data and combined with physical principles (such as solar radiation, shading effect, and temperature effect), the solar power physics model calculates the sample output power of the solar cells under these conditions.

[0063] Then, the difference between the sample output power calculated by the solar power physics model and the actual output power of the solar cells is compared. Based on this error, a corresponding solar panel output power error label is assigned to each sample operating environment state data. This label represents the deviation between the model prediction and the actual value.

[0064] Furthermore, the environmental status data of all samples labeled with solar panel output power error are compiled into a training sample set. Then, the deep learning model is trained using this training sample set. In this invention, the deep learning model is built based on an improved Blur informer model. In the improved Blur informer model, a low-pass filter is placed between the max pooling layer and the downsampling layer. The low-pass filter smooths the input signal, reducing the impact of high-frequency noise, thereby potentially improving the model's stability and prediction accuracy.

[0065] Through training, the deep learning model learns the complex relationship between the sample operating environment state data and the solar cell output power, thus obtaining a solar power prediction model capable of predicting the error in solar cell output power. The deep learning model is used here to predict the output error of the solar cell. Since the prediction of the physical model may contain errors, it is necessary to use the output error predicted by the deep learning model to adjust the initial output power of the physical model. Through this adjustment, a prediction value closer to the actual output power can be obtained, with higher accuracy and reliability.

[0066] This invention further optimizes accuracy using the Blur Informer model, building upon the existing physical model. Physical models are typically based on a series of assumptions and simplified equations, capturing some fundamental physical phenomena. However, real-world situations are often far more complex than these physical models, especially when considering multimodal characteristics and subtle shading effects. Deep learning models can learn and capture these complex nonlinear relationships, thereby improving prediction accuracy. In this invention, the deep learning model adapts to changing environments and conditions through continuous training and adjustment. It can quickly adjust to weather changes, seasonal variations, and other factors that may affect solar power generation, maintaining prediction accuracy. The improved Blur Informer model in this invention can handle a large number of input features, including sensor data, environmental conditions, and historical data, to assess the impact of various factors on solar power generation. Furthermore, the deep learning model corrects errors in the physical model, particularly for specific situations or subtle influences (such as shading effects and attitude changes) that the physical model failed to fully capture. This invention addresses the poor translation invariance of self-attention distillation in traditional Informers by modifying the max-pooling layer in the encoder of the standard Informer to a fuzzy pooling layer. Specifically, to combat aliasing, this invention adds a low-pass filter between the dense evaluation in max pooling and the subsequent simple downsampling, enabling a single-layer stack in the Blur informer to achieve anti-aliasing.

[0067] The solar power generation prediction system for aircraft provided by the present invention is described below. The solar power generation prediction system for aircraft described below can be referred to in correspondence with the solar power generation prediction method for aircraft described above.

[0068] Figure 2 This is a schematic diagram of the structure of the solar power generation prediction system for aircraft provided by the present invention, as shown below. Figure 2As shown, this invention provides a solar power generation prediction system for aircraft, including a data acquisition module 201 and a solar power generation prediction module 202. The data acquisition module 201 acquires target operating environment state data corresponding to the target aircraft. This target operating environment state data includes at least Earth-Sun distance data, Earth orbital dynamics parameters, aircraft shading parameters, ambient temperature parameters, and aircraft sensor data. The aircraft shading parameters represent the shading range of solar cells by components of the target aircraft. The solar power generation prediction module 202 inputs the target operating environment state data into the solar power physical model and the solar power prediction model in the solar power generation prediction integrated model, respectively, to obtain the solar power generation prediction integrated model. The model outputs the predicted output power of the solar array corresponding to the target aircraft; wherein, the solar power physical model is used to obtain the first solar cell output power based on total solar radiation, effective solar cell shading area, and temperature effect, wherein the total solar radiation, the effective solar cell shading area, and the temperature effect are calculated based on the target operating environment state data; the solar power prediction model is trained based on a deep learning model and is used to predict the second solar cell output power based on the input target operating environment state data; the solar power generation prediction comprehensive model is used to adjust the first solar cell output power based on the second solar cell output power to obtain the predicted output power of the solar array.

[0069] The solar power generation prediction system for aircraft provided by this invention acquires the Earth-Sun distance, Earth orbit dynamics parameters, aircraft shading parameters, ambient temperature, and aircraft sensor data of the target near-space aircraft. These data are then input into two sub-models in a comprehensive solar power generation prediction model. The comprehensive model then combines the outputs of the two sub-models and adjusts the first output power of the physical model by using the second output power output by the deep learning model, thereby obtaining a more accurate predicted output power of the solar cell array.

[0070] The system provided in this embodiment of the invention is used to execute the above-described method embodiments. For specific processes and details, please refer to the above embodiments, which will not be repeated here.

[0071] Figure 3 This is a schematic diagram of the structure of the electronic device provided by the present invention, such as... Figure 3As shown, the electronic device may include: a processor 301, a communications interface 302, a memory 303, and a communication bus 304, wherein the processor 301, communications interface 302, and memory 303 communicate with each other via the communication bus 304. The processor 301 can call logical instructions in the memory 303 to execute a solar power generation prediction method for an aircraft. This method includes: acquiring target operating environment state data corresponding to the target aircraft, wherein the target operating environment state data includes at least Earth-Sun distance data, Earth orbit dynamic parameters, aircraft shading parameters, ambient temperature parameters, and aircraft sensor data; the aircraft shading parameters represent the shading range of solar cells by components of the target aircraft; the target aircraft is a near-space aircraft; and inputting the target operating environment state data into the solar power physical model and the solar power prediction model in the solar power generation prediction integrated model, respectively, to obtain the target aircraft's solar power prediction prediction data output by the solar power generation prediction integrated model. The solar array output power corresponding to the target aircraft is predicted; wherein, the solar power physical model is used to obtain the first solar cell output power based on the total solar radiation, the effective area of ​​solar cell shading, and the temperature effect, wherein the total solar radiation, the effective area of ​​solar cell shading, and the temperature effect are calculated based on the target operating environment state data; the solar power prediction model is trained based on a deep learning model and is used to predict the second solar cell output power based on the input target operating environment state data; the solar power generation prediction comprehensive model is used to adjust the first solar cell output power based on the second solar cell output power to obtain the predicted output power of the solar array.

[0072] Furthermore, the logical instructions in the aforementioned memory 303 can be implemented as software functional units and, when sold or used as independent products, can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, essentially, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0073] On the other hand, the present invention also provides a computer program product, the computer program product comprising a computer program stored on a non-transitory computer-readable storage medium, the computer program comprising program instructions, which, when executed by a computer, enable the computer to execute the solar power generation prediction method for aircraft provided by the above methods, the method comprising: acquiring target operating environment state data corresponding to the target aircraft, wherein the target operating environment state data includes at least Earth-Sun distance data, Earth orbit dynamic parameters, aircraft shading parameters, ambient temperature parameters, and aircraft sensor data; the aircraft shading parameters represent the shading range of solar cells by components of the target aircraft; the target aircraft is a near-space aircraft; and inputting the target operating environment state data into the solar power object of the solar power generation prediction integrated model. In the physical model and the solar power prediction model, the predicted output power of the solar cell array corresponding to the target aircraft is obtained by the integrated solar power prediction model. Specifically, the physical solar power model is used to obtain the first solar cell output power based on total solar radiation, the effective area of ​​solar cell shading, and the temperature effect, which are calculated based on the target operating environment state data. The solar power prediction model is trained based on a deep learning model and is used to predict the second solar cell output power based on the input target operating environment state data. The integrated solar power prediction model is used to adjust the first solar cell output power based on the second solar cell output power to obtain the predicted output power of the solar cell array.

[0074] In another aspect, the present invention also provides a non-transitory computer-readable storage medium storing a computer program thereon, which, when executed by a processor, implements the solar power generation prediction method for aircraft provided in the above embodiments. The method includes: acquiring target operating environment state data corresponding to the target aircraft, wherein the target operating environment state data includes at least Earth-Sun distance data, Earth orbital dynamics parameters, aircraft shading parameters, ambient temperature parameters, and aircraft sensor data; the aircraft shading parameters represent the shading range of solar cells by components of the target aircraft; the target aircraft is a near-space aircraft; and inputting the target operating environment state data into the solar power physical model and the solar power prediction model in the solar power generation prediction integrated model, respectively, to obtain a result from... The solar power prediction model outputs the predicted power of the solar cell array corresponding to the target aircraft. The solar power physical model is used to obtain the first solar cell output power based on total solar radiation, the effective area of ​​solar cell shading, and the temperature effect. The total solar radiation, the effective area of ​​solar cell shading, and the temperature effect are calculated based on the target operating environment status data. The solar power prediction model is trained based on a deep learning model and is used to predict the second solar cell output power based on the input target operating environment status data. The solar power prediction model is used to adjust the first solar cell output power based on the second solar cell output power to obtain the predicted output power of the solar cell array.

[0075] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs. Those skilled in the art can understand and implement this without any creative effort.

[0076] Through the above description of the embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus necessary general-purpose hardware platforms, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solutions, in essence or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods described in the various embodiments or some parts of the embodiments.

[0077] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for predicting solar power generation for aircraft, characterized in that, include: Acquire target operating environment status data corresponding to the target spacecraft, wherein the target operating environment status data includes at least Earth-Sun distance data, Earth orbital dynamics parameters, spacecraft shading parameters, ambient temperature parameters, and spacecraft sensor data; the spacecraft shading parameters represent the shading range of the solar cells by the components of the target spacecraft; the target spacecraft is a near-space spacecraft; The target operating environment state data is input into the solar power physical model and the solar power prediction model in the solar power prediction integrated model, respectively, to obtain the predicted output power of the solar cell array corresponding to the target aircraft, output by the solar power prediction integrated model. Specifically, the solar power physical model is used to obtain the first solar cell output power based on total solar radiation, the effective area of ​​solar cell shading, and the temperature effect, which are calculated based on the target operating environment state data. The solar power prediction model is trained based on a deep learning model and is used to predict the second solar cell output power based on the input target operating environment state data. The solar power prediction integrated model is used to adjust the first solar cell output power based on the second solar cell output power to obtain the predicted output power of the solar cell array.

2. The solar power generation prediction method for aircraft according to claim 1, characterized in that, The solar power physical model includes a solar radiation model, a shadowing model, a temperature effect model, and an output power model, wherein: The solar radiation model is used to calculate the total solar radiation corresponding to the location of the target spacecraft at the current time, based on the Earth-Sun distance data, the Earth orbital dynamics parameters, the spacecraft's spatial position, and the current time. The shadow model is used to calculate the shadow occlusion area of ​​the aircraft balloon and the shadow occlusion area of ​​the aircraft flange according to the aircraft occlusion parameters, and to obtain the effective solar cell occlusion area of ​​the target aircraft at the current time based on the shadow occlusion area of ​​the aircraft balloon and the shadow occlusion area of ​​the aircraft flange. The effective solar cell occlusion area represents the area corresponding to the solar cell area on the aircraft that is not occluded by the aircraft balloon and the aircraft flange. The temperature effect model is used to determine the real-time temperature data and reference temperature data corresponding to the location of the target aircraft at the current time based on the ambient temperature parameters, and to calculate the temperature effect based on the real-time temperature data and the reference temperature data. The output power model is used to calculate the output power of the first solar cell based on the total solar radiation, the effective shading area of ​​the solar cell, and the temperature effect.

3. The solar power generation prediction method for aircraft according to claim 2, characterized in that, The calculation formula for the solar radiation model is as follows: ； ； ； ； ； ； in, This represents the total solar radiation. This indicates the intensity of direct solar radiation. Indicates the intensity of reflected radiation. Indicates the solar altitude angle. Indicates sunrise time. Indicates sunset time. Indicates reflectance on sunny days. This represents the corrected extraterrestrial solar radiation. Indicates atmospheric transmittance. Indicates air quality, This indicates the atmospheric pressure at the location of the aircraft. This represents the atmospheric pressure at sea level. Indicates extraterrestrial solar radiation; This represents the eccentricity factor calculated from the Earth-Sun distance data.

4. The solar power generation prediction method for aircraft according to claim 3, characterized in that, The calculation formula for the shadow model is as follows: ； ； ； in, This indicates the area of ​​the shadow cast by the aircraft balloon. This represents the area directly exposed to sunlight. Indicates the length of the solar cell array. Indicates the width of the solar cell array. This indicates the area of ​​the aircraft flange that is shaded. Indicates the flange shadow angle. Indicates the first critical angle of shadow. Indicates the second critical angle of shadow. Indicates the length of the aircraft flange. Indicates the width of the aircraft flange. This indicates the effective shading area of ​​the solar cell.

5. The solar power generation prediction method for aircraft according to claim 4, characterized in that, The calculation formula for the temperature effect model is as follows: ； in, This refers to the temperature effect. This indicates the temperature effect corresponding to the reference temperature data. Indicates the temperature coefficient. This refers to the real-time temperature data. This refers to the reference temperature data; The calculation formula for the output power model is as follows: ； in, This indicates the output power of the first solar cell. This indicates the projection rate of the outer layer of the solar array.

6. The method for predicting solar power generation for aircraft according to any one of claims 1 to 5, characterized in that, The solar power prediction model is trained through the following steps: Acquire the sample operating environment status data and the actual output power of the solar panel cell corresponding to the sample operating environment status data; The sample operating environment status data is input into the solar power physical model to obtain the solar cell sample output power output by the solar power physical model; Based on the error between the actual output power of the solar panel cell and the output power of the solar cell sample, a solar panel cell output power error label corresponding to the sample operating environment status data label is constructed. A training sample set is constructed based on the sample operating environment status data labeled with solar panel cell output power error tags; The deep learning model is trained based on the training sample set to obtain the solar power prediction model. The deep learning model is constructed based on the improved Blur informer model, in which a low-pass filter is set between the max pooling layer and the downsampling layer.

7. A solar power generation prediction system for aircraft, characterized in that, include: The data acquisition module is used to acquire target operating environment status data corresponding to the target spacecraft. The target operating environment status data includes at least Earth-Sun distance data, Earth orbit dynamics parameters, spacecraft shading parameters, ambient temperature parameters, and spacecraft sensor data. The spacecraft shading parameters represent the shading range of the solar cells by the target spacecraft's components. The solar power prediction module is used to input the target operating environment status data into the solar power physical model and the solar power prediction model in the solar power prediction integrated model, respectively, to obtain the predicted output power of the solar cell array corresponding to the target aircraft output by the solar power prediction integrated model; The solar power physics model is used to obtain the output power of the first solar cell based on total solar radiation, effective shading area of ​​the solar cell, and temperature effect. The total solar radiation, effective shading area of ​​the solar cell, and temperature effect are calculated based on the target operating environment state data. The solar power prediction model is trained based on a deep learning model and is used to predict the output power of the second solar cell based on the input target operating environment state data. The solar power generation prediction comprehensive model is used to adjust the output power of the first solar cell based on the output power of the second solar cell to obtain the predicted output power of the solar cell array.

8. An electronic device comprising a memory, a processor, and a computer program stored in the memory and running on the processor, characterized in that, When the processor executes the computer program, it implements the solar power generation prediction method for aircraft as described in any one of claims 1 to 6.

9. A non-transitory computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the solar power generation prediction method for aircraft as described in any one of claims 1 to 6.

10. A computer program product, comprising a computer program, characterized in that, When the computer program is executed by the processor, it implements the solar power generation prediction method for aircraft as described in any one of claims 1 to 6.

Images