Solar field cloud shading monitoring and control method and system based on multi-sensor data fusion

Abstract

The present application relates to a kind of solar field cloud shelter monitoring and control method, system based on multisensor data fusion.The method includes: the low-density scattered solar radiation sensor network and high-density overall solar radiation sensor network are cooperatively collected basic data in deployment, combined with adaptive filtering technique to filter noise interference, then based on the inverse deduction direct solar radiation intensity of radiation balance equation;While aiming at shelter area, non-shelter area uses differentiated interpolation and extrapolation method;Through convolutional neural network, cloud layer features are extracted, cloud movement trajectory and shelter range are predicted;Introduce geographic information system data, according to the projection length and shape of topographic relief correction cloud shadow;While fusing meteorological satellite data and cloud height data to construct cloud layer three-dimensional model;Based on direct solar radiation intensity distribution map and cloud shelter trajectory prediction result, the working state of heliostat in shelter area is dynamically adjusted, to avoid invalid operation and energy waste.

Description

Technical Field

[0001] This invention relates to the field of solar thermal power generation monitoring and control technology, and in particular to a method and system for monitoring and controlling cloud shading in solar fields based on multi-sensor data fusion. Background Technology

[0002] Solar thermal power generation systems are an important renewable energy utilization technology. Their core principle involves reflecting and concentrating sunlight onto a receiver located at the top of a tower using heliostats, converting light energy into high-temperature heat energy to drive a steam turbine generator to produce electricity. In solar thermal power generation systems, the intensity of direct solar radiation is a key factor affecting energy capture and conversion efficiency. However, dynamic cloud cover is a major challenge for solar power plant operation, often causing rapid and drastic fluctuations in the direct radiation energy received by the heliostats. This not only leads to unstable receiver operating temperatures but also severely impacts the overall system's power generation efficiency and output stability. To address the uncertainties introduced by clouds, cloud cover phenomena can be monitored, predicted, and optimized. Commonly used methods include acquiring cloud images using ground-based or sky-based imaging equipment to generate cloud mapping maps, and combining this with measurement data from a sparsely distributed radiation sensor network to estimate and interpolate cloud cover conditions.

[0003] However, traditional methods for dealing with cloud shading are flawed. Data collected by a single sensor is susceptible to environmental interference, resulting in significant errors and insufficient accuracy in estimating direct solar radiation. Furthermore, the generated cloud maps cannot accurately define cloud shadow boundaries, leading to deviations in cloud-shading area delineation. It is difficult to predict the development trend of cloud shading in advance, leaving insufficient time for heliostat deployment adjustments. The methods also fail to adequately consider complex terrain features, exhibiting poor adaptability in mountainous and other special terrain environments, further reducing the accuracy of direct solar radiation intensity estimation. Additionally, the lack of real-time control strategies linked to cloud shading monitoring results prevents dynamic adjustments to heliostat operation to mitigate cloud shading losses. The methods also lack the ability to handle complex cloud scenarios such as multi-layered cloud superposition, making it difficult to accurately assess the impact of cloud shading on radiation energy in these scenarios. Finally, the high deployment cost of sensor networks significantly restricts their widespread application in large-scale solar thermal power plants.

[0004] Therefore, traditional methods for handling cloud shading in solar fields often suffer from problems such as insufficient accuracy in monitoring cloud shading and severe energy loss due to insufficient cloud height information and motion prediction, poor algorithm adaptability, and lack of real-time control strategies. Summary of the Invention

[0005] In order to solve the above-mentioned technical problems, a method and system for monitoring and controlling cloud shading of solar fields based on multi-sensor data fusion is provided. This method and system can improve the accuracy of direct radiation estimation and the estimation accuracy under complex terrain, and effectively avoid energy loss caused by cloud shading.

[0006] A method for monitoring and controlling cloud shading in solar fields based on multi-sensor data fusion, the method comprising:

[0007] A low-density scattered solar radiation sensor network and a high-density overall solar radiation sensor network are deployed inside the solar field to acquire scattered solar radiation data and overall solar radiation data in different regions of the solar field.

[0008] Adaptive filtering technology is used to process the scattered solar radiation data and the total solar radiation data. Based on the processed solar radiation data, the direct solar radiation intensity is estimated using the radiation balance equation.

[0009] Ground and sky image data and cloud height data are acquired by fisheye camera and lidar, and convolutional neural network is used to analyze the ground and sky image data and cloud height data to predict cloud obscuring trajectory.

[0010] Based on the cloud obscuration trajectory, the monitoring points are divided into obscured and unobscured areas. The direct solar radiation intensity is spatially interpolated using the overall solar radiation interpolation and the scattered solar radiation extrapolation methods to obtain a direct radiation intensity distribution map. Based on the direct radiation intensity distribution map, the working status of the heliostats in the obscured area is adjusted.

[0011] By introducing geographic information system data, the projected length and shape of the cloud shadow on the terrain are calculated and corrected based on the cloud shading trajectory and terrain undulation. By fusing meteorological satellite data and cloud height data, a three-dimensional cloud model is constructed, and the impact of cloud height changes on the intensity of direct solar radiation is analyzed based on the three-dimensional cloud model, so as to realize cloud shading monitoring and control.

[0012] In one embodiment, a low-density scattered solar radiation sensor network and a high-density overall solar radiation sensor network are deployed within the solar field to acquire scattered solar radiation data and overall solar radiation data for different regions of the solar field, including:

[0013] Based on the size of the solar field, determine the number and location of low-density scattered solar radiation sensors and high-density overall solar radiation sensors.

[0014] Based on the number and location of the aforementioned sensors, a low-density scattered solar radiation sensor network is installed to measure scattered solar radiation data in different areas of the solar field.

[0015] Based on the number and location of the sensors, a high-density total solar radiation sensor network is installed to measure total solar radiation data in different areas of the solar field.

[0016] In one embodiment, adaptive filtering technology is used to process the scattered solar radiation data and the total solar radiation data. Based on the processed solar radiation data, the direct solar radiation intensity is estimated using the radiation balance equation, including:

[0017] Based on the scattered solar radiation data and the total solar radiation data, the intensity of direct solar radiation is estimated using the radiation balance equation.

[0018] Adaptive filtering technology is used to filter sensor noise data, and a real-time data calibration algorithm is introduced to dynamically optimize sensor data weights, thereby completing the processing of the scattered solar radiation data and the total solar radiation data.

[0019] The adaptive filtering technique is selected from one of the following: Kalman filtering algorithm, wavelet transform denoising algorithm, and median filtering algorithm; the real-time data calibration algorithm is selected from one of the following: least squares method, Bayesian estimation, and maximum entropy algorithm.

[0020] In one embodiment, a convolutional neural network is used to analyze the ground and sky image data and cloud height data to predict cloud obscuration trajectories, including:

[0021] The ground and sky image data and cloud height data are analyzed using a convolutional neural network to extract cloud features;

[0022] Based on the cloud characteristics, cloud boundaries and cloud shadow projections are determined using the cloud height data.

[0023] Based on time series analysis, predict the speed, direction, and occlusion of cloud movement;

[0024] Based on the cloud boundary, cloud shadow projection, cloud movement speed, direction, and occlusion status, the cloud occlusion trajectory is predicted.

[0025] In one embodiment, the spatial interpolation difference method is selected from inverse distance weighted interpolation and Kriging interpolation; the extrapolation method is selected from linear extrapolation and polynomial extrapolation.

[0026] In one embodiment, adjusting the operating state of the heliostat within the shielded area according to the direct radiation intensity distribution map includes:

[0027] The cluster of heliostats within the shielded area is determined based on the direct radiation intensity distribution map.

[0028] The aiming logic of the heliostat cluster within the shielded area is adjusted in real time, and the heliostat cluster within the shielded area is divided into low-priority heliostats and high-priority heliostats based on the aiming logic.

[0029] Switch the low-priority heliostat to standby mode and switch the high-priority heliostat to standby heat storage and reflection mode.

[0030] Once the clouds have dispersed, the high-priority heliostats and low-priority heliostats will be switched back to their original positions based on the direct solar radiation intensity measurement results.

[0031] In one embodiment, geographic information system data is introduced, and the length and shape of the cloud shadow projection on the terrain are calculated and corrected based on the cloud shading trajectory and terrain undulation, including:

[0032] Geographic information system data is introduced to obtain the topographic information of the solar field;

[0033] Based on the cloud occlusion trajectory, the terrain undulation is determined according to the terrain information, and the length and shape of the cloud shadow projection on the terrain are corrected by the terrain correction algorithm.

[0034] The terrain correction algorithm is either a ray tracing algorithm or a ray tracing algorithm.

[0035] In one embodiment, the method further includes:

[0036] The deployment of low-density scattered solar radiation sensor network and high-density overall solar radiation sensor network was periodically calibrated on-site using a UAV equipped with a standard radiometer, and the calibration results were obtained.

[0037] Adjust the parameters of each sensor based on the calibration results.

[0038] In one embodiment, the method further includes:

[0039] Determine whether it is necessary to continue monitoring the solar field cloud shading. If so, collect scattered solar radiation data and total solar radiation data again, estimate the direct solar radiation intensity, predict the cloud shading trajectory, adjust the working status of the heliostat in the shading area, and form a closed-loop monitoring and control system.

[0040] If continued monitoring of solar field cloud shading is no longer required, then the process ends.

[0041] A solar field cloud shading monitoring and control system based on multi-sensor data fusion, the system comprising:

[0042] The data acquisition module is used to deploy a low-density scattered solar radiation sensor network and a high-density overall solar radiation sensor network inside the solar field to acquire scattered solar radiation data and overall solar radiation data in different areas of the solar field.

[0043] The direct solar radiation intensity estimation module is used to process the scattered solar radiation data and the total solar radiation data using adaptive filtering technology, and to estimate the direct solar radiation intensity based on the processed solar radiation data through the radiation balance equation.

[0044] The cloud obscuration trajectory prediction module is used to acquire ground and sky image data and cloud height data through fisheye camera and lidar, and to analyze the ground and sky image data and cloud height data using convolutional neural network to predict cloud obscuration trajectory.

[0045] The heliostat adjustment module is used to divide the monitoring point into a shaded area and an unshaded area according to the cloud shading trajectory, and to perform spatial interpolation of the direct solar radiation intensity using the overall solar radiation interpolation and the scattered solar radiation extrapolation method respectively to obtain a direct radiation intensity distribution map. Based on the direct radiation intensity distribution map, the module adjusts the working state of the heliostat in the shaded area.

[0046] The analysis and monitoring module is used to import data from the geographic information system, calculate and correct the projection length and shape of the cloud shadow on the terrain based on the cloud shading trajectory and the terrain undulation; integrate meteorological satellite data and cloud height data to construct a three-dimensional cloud model, and analyze the impact of cloud height changes on the intensity of direct solar radiation based on the three-dimensional cloud model, so as to realize cloud shading monitoring and control.

[0047] The aforementioned method and system for monitoring and controlling solar field cloud shading based on multi-sensor data fusion, by deploying a low-density scattered solar radiation sensor network and a high-density overall solar radiation sensor network to collaboratively collect basic data, combined with adaptive filtering technology to filter noise interference, and then deriving the direct solar radiation intensity based on the radiation balance equation, solves the problem of large errors in single-sensor data. Simultaneously, differentiated interpolation and extrapolation methods are used for shaded and unshaded areas to further improve the accuracy of direct solar radiation intensity estimation. Cloud features are extracted through convolutional neural networks to achieve refined identification of cloud boundaries and accurate prediction of cloud movement trajectories and shading ranges. Geographic information system data is introduced to correct the projection length and shape of cloud shadows according to terrain undulations, solving the problem of large shadow estimation deviations in complex terrain. Furthermore, a three-dimensional cloud model is constructed by integrating meteorological satellite data and cloud height data, enabling analysis of the impact of cloud height changes on direct solar radiation intensity. Based on the direct solar radiation intensity distribution map and cloud shading trajectory prediction results, the working status of heliostats within the shading area is dynamically adjusted to avoid ineffective operation and energy waste. Attached Figure Description

[0048] Figure 1 This is an application environment diagram of a solar field cloud shading monitoring and control method based on multi-sensor data fusion in one embodiment.

[0049] Figure 2 This is a flowchart illustrating a method for monitoring and controlling cloud shading in a solar field based on multi-sensor data fusion in one embodiment.

[0050] Figure 3 This is a block diagram of a solar field cloud shading monitoring and control system based on multi-sensor data fusion in one embodiment.

[0051] Figure 4 This is an internal structural diagram of a computer device in one embodiment. Detailed Implementation

[0052] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0053] The solar field cloud shading monitoring and control method based on multi-sensor data fusion provided in this application can be applied to, for example... Figure 1 The application environment shown. For example... Figure 1As shown, the application environment includes a computer device 110 and a low-density scattered solar radiation sensor network 120, a high-density overall solar radiation sensor network 130, a fisheye camera 140, and a lidar 150 connected to the computer device 110. The low-density scattered solar radiation sensor network 120 and the high-density overall solar radiation sensor network 130 are deployed within the solar field to acquire scattered solar radiation data and overall solar radiation data from different areas of the solar field. The computer device 110 can use adaptive filtering technology to process the scattered solar radiation data and overall solar radiation data, and estimate the direct solar radiation intensity based on the processed solar radiation data using the radiation balance equation. Ground and sky image data and cloud height data are acquired through the fisheye camera 140 and lidar 150. The computer device 110 can use a convolutional neural network to analyze the ground and sky image data and cloud height data to predict cloud obscuring trajectories. The system 110 can divide monitoring points into shaded and unshaded areas based on cloud shading trajectories. It then uses both overall solar radiation interpolation and diffuse solar radiation extrapolation methods to spatially interpolate the direct solar radiation intensity, obtaining a direct radiation intensity distribution map. Based on this map, it adjusts the operating status of heliostats within the shaded areas. The computer equipment 110 can import geographic information system data, calculate and correct the projection length and shape of cloud shadows on the terrain based on the cloud shading trajectory and topographic relief. It integrates meteorological satellite data and cloud height data to construct a three-dimensional cloud model, and analyzes the impact of cloud height changes on direct solar radiation intensity based on this model, thus achieving cloud shading monitoring and control. The computer equipment 110 can be, but is not limited to, various personal computers, laptops, robots, unmanned aerial vehicles, tablets, and other similar devices.

[0054] In one embodiment, such as Figure 2 As shown, a method for monitoring and controlling cloud shading in solar fields based on multi-sensor data fusion is provided, including the following steps:

[0055] Step 202: Deploy a low-density scattered solar radiation sensor network and a high-density overall solar radiation sensor network inside the solar field to acquire scattered solar radiation data and overall solar radiation data for different regions of the solar field.

[0056] To achieve comprehensive, real-time, and accurate acquisition of scattered solar radiation data and total solar radiation data in different regions of the solar field, it is necessary to deploy a low-density scattered solar radiation sensor network and a high-density total solar radiation sensor network within the solar field to measure the scattered solar radiation and total solar radiation data in different regions of the solar field.

[0057] Specifically, in one embodiment, a method for monitoring and controlling cloud shading of a solar field based on multi-sensor data fusion may further include the process of collecting solar radiation data from different regions. The specific process includes: determining the number and location of low-density scattered solar radiation sensors and high-density overall solar radiation sensors according to the size of the solar field; installing a low-density scattered solar radiation sensor network based on the number and location of the sensors to measure scattered solar radiation data from different regions of the solar field; and installing a high-density overall solar radiation sensor network based on the number and location of the sensors to measure overall solar radiation data from different regions of the solar field.

[0058] For low-density scattered solar radiation sensor networks, the number of sensors is 10-30, depending on the size of the solar field. For example, 10-15 sensors for small fields, 15-25 sensors for medium fields, and 20-30 sensors for large fields. The sensor spacing is controlled at 500-800 meters. Sensors are preferentially placed at the edges of the solar field and at turning points of terrain undulations, ensuring coverage of different scattered radiation environments within the field. For high-density overall solar radiation sensor networks, the number of sensors is 100-300, with 100-150 sensors for small fields, 150-250 sensors for medium fields, and 200-300 sensors for large fields. The sensor spacing is controlled at 100-300 meters. Sensors are uniformly arranged in a grid pattern to cover the entire heliostat cluster area of ​​the solar field, ensuring that the spatial resolution of the overall solar radiation data meets the requirements for subsequent zonal estimation.

[0059] All sensors can be mounted on brackets to ensure a stable installation and prevent the sensors from shaking due to wind or other factors. The sensor's detection surface must be horizontal and facing upwards, and there should be no tall buildings, trees, or other obstructions nearby to ensure unobstructed reception of solar radiation.

[0060] In a low-density scattered solar radiation sensor network, each sensor collects scattered solar radiation data (DH) of its area in real time. The collected parameters include scattered radiation intensity and collection timestamp. The collected data is uploaded to a computer device in real time via wired or wireless communication modules. In a high-density total solar radiation sensor network, each sensor collects total solar radiation data (GH) of its area in real time. The collected parameters include total radiation intensity and collection timestamp. The collected data is uploaded to a computer device synchronously.

[0061] Step 204: Adaptive filtering technology is used to process the scattered solar radiation data and the overall solar radiation data. Based on the processed solar radiation data, the direct solar radiation intensity is estimated using the radiation balance equation.

[0062] After receiving the collected scattered solar radiation data and total solar radiation data, the computer equipment can estimate the direct solar radiation intensity through the radiation balance equation and use adaptive filtering technology to process the sensor data, thereby improving the accuracy and robustness of the direct radiation estimation.

[0063] In one embodiment, a method for monitoring and controlling cloud shading of a solar field based on multi-sensor data fusion may further include a process for estimating the intensity of direct solar radiation. Specifically, this process includes: estimating the intensity of direct solar radiation based on scattered solar radiation data and total solar radiation data using the radiation balance equation; filtering sensor noise data using adaptive filtering technology and introducing a real-time data calibration algorithm to dynamically optimize sensor data weights, thereby completing the processing of scattered solar radiation data and total solar radiation data; wherein the adaptive filtering technology is selected from one of the following algorithms: Kalman filtering, wavelet transform denoising, and median filtering; and the real-time data calibration algorithm is selected from one of the following algorithms: least squares, Bayesian estimation, and maximum entropy.

[0064] Computer equipment can estimate the direct solar radiation intensity DNR using scattered solar radiation and total solar radiation data based on the radiation balance equation GH=DH+DNR×cosθ0. It also employs adaptive filtering technology to process noise data in the sensor network, improving the robustness of DNR estimation. Furthermore, it introduces a real-time data calibration algorithm to dynamically optimize the weights of different sensor data, with a weight range of 0.1-0.9, thereby reducing error accumulation.

[0065] Step 206: Obtain ground and sky image data and cloud height data through fisheye camera and lidar, and use convolutional neural network to analyze the ground and sky image data and cloud height data to predict cloud obscuring trajectory.

[0066] Fisheye cameras and lidar can also be installed inside the solar field to acquire ground and sky image data, cloud height data, etc. Among them, fisheye cameras and lidar can be deployed in a multi-point distributed manner, with priority given to the highest points of the solar field and the edges of the core area of ​​the heliostat cluster, to ensure that the detection range of the equipment can cover the entire solar field and the surrounding sky area, avoiding interference from terrain obstruction or reflection from the heliostat itself on data acquisition.

[0067] Once the fisheye camera is activated, it can continuously collect images of the ground and sky. Each frame contains information such as cloud formations, sun position, and distribution of ground heliostats. The collected data is uploaded to the computer in real time. The lidar can be activated synchronously with the fisheye camera. By emitting a laser beam into the sky and receiving the reflected signal, it calculates the vertical distance between the cloud layer and the device, generates a cloud height dataset, which includes cloud height values, detection point coordinates, and timestamps, and uploads it to the computer in real time.

[0068] In one embodiment, a method for monitoring and controlling cloud shading in a solar field based on multi-sensor data fusion may further include a process for predicting cloud shading trajectories. Specifically, this process includes: analyzing ground and sky image data and cloud height data using a convolutional neural network to extract cloud features; determining cloud boundaries and cloud shadow projections based on cloud height data according to cloud features; predicting the speed, direction, and shading status of cloud movement based on time series analysis; and predicting the cloud shading trajectory based on cloud boundaries, cloud shadow projections, and the speed, direction, and shading status of cloud movement.

[0069] Computer equipment can use convolutional neural networks to analyze sequences of ground and sky images and extract cloud features; combine cloud height data to refine cloud boundaries and shadow projections; and predict the speed, direction, and occlusion trajectory of cloud movement based on time series analysis.

[0070] Among them, the convolutional neural network can use ResNet-50 or Inception-v3 convolutional neural network as the core model. It has a deep network structure and powerful feature extraction capabilities, and can effectively capture the morphological change patterns and highly correlated features of clouds.

[0071] Computer equipment can input data collected by fisheye cameras and LiDAR into a trained convolutional neural network. Through convolutional and pooling layers, it gradually extracts deep features of clouds, such as shape, texture, movement trend, and height changes. Based on the extracted cloud features and combined with time series analysis methods, it outputs the cloud's movement speed, direction, and cloud obscuring trajectory within the solar field range for the next 5-30 minutes. Specifically, this is represented by the coordinate range, coverage area, and duration of obscuring at different time points. Finally, the prediction results can be stored in the computer equipment as a time-space correlated dataset and visualized as a dynamic cloud obscuring trajectory map. This provides a basis for subsequent decision-making regarding the division of obscuring and unobscuring areas and the adjustment of heliostat working status, enabling accurate acquisition and in-depth mining of ground and sky image data and cloud height data, thereby improving the accuracy and timeliness of cloud obscuring trajectory prediction.

[0072] Step 208: Divide the monitoring points into shaded and unshaded areas according to the cloud shading trajectory. Use the overall solar radiation interpolation and the scattered solar radiation extrapolation method to spatially interpolate the direct solar radiation intensity to obtain the direct radiation intensity distribution map. Adjust the working status of the heliostat in the shaded area according to the direct radiation intensity distribution map.

[0073] Computer equipment can classify direct radiation distribution points into shaded and unshaded areas based on cloud cover status, and estimate direct radiation intensity using both total solar radiation interpolation and diffuse solar radiation extrapolation. Specifically, the computer equipment can divide monitoring points into shaded and unshaded areas based on cloud cover status; for shaded areas, the DNR is estimated using interpolation methods from nearby total solar radiation sensors; for unshaded areas, the DNR is estimated by fusing diffuse solar radiation sensor data through extrapolation.

[0074] The computer equipment can pre-divide the entire solar field into monitoring points using a grid, with each monitoring point corresponding to unique coordinate information, and associate it with nearby low-density scattered solar radiation sensors and high-density total solar radiation sensors. Then, it can extract key information from the cloud obscuring trajectory prediction results, including the coordinate range, coverage boundary, and duration of the obscuring area at different time points. The coordinates of each monitoring point are spatiotemporally matched with the coordinates of the obscuring area to determine whether each monitoring point will be covered by clouds in the future prediction period. If a monitoring point is covered by a cloud obscuring trajectory for any consecutive period within the next 5-30 minutes, and the initial predicted DNR value for that period is lower than a set threshold, then the monitoring point is classified as an obscuring area. If a monitoring point is not covered by a cloud obscuring trajectory, or is briefly covered but for less than 1 minute, and the initial predicted DNR value is higher than a set threshold, then it is classified as an unobscuring area. Finally, the computer equipment can summarize the partitioning results of all monitoring points to generate a spatial distribution map of the solar field's obscuring and unobscuring areas.

[0075] In one embodiment, the spatial interpolation difference method is selected from inverse distance weighted interpolation and Kriging interpolation; the extrapolation method is selected from linear extrapolation and polynomial extrapolation.

[0076] In other words, cloud cover in the shaded area leads to a significant attenuation of direct radiation and an increase in the proportion of scattered radiation. Therefore, total solar radiation (GH) data better reflects the total radiation characteristics of the region. Thus, GH data collected by high-density total solar radiation sensors in the shaded area and its surrounding areas are selected as the basis for interpolation. Spatial interpolation is performed using either inverse distance weighted interpolation or Kriging interpolation. In inverse distance weighted interpolation, the distance from the monitoring point to each sensor node is used as the weight, with closer distances having a larger weight. The weighted average is used to calculate the estimated GH value for the monitoring point, which is then combined with empirical values ​​of the proportion of scattered radiation in the area to back-calculate the estimated DNR value. Kriging interpolation, based on regional variable theory, analyzes the spatial correlation of GH data through a semi-variogram, constructs an interpolation model, predicts the GH data for the monitoring point, and then converts it into the DNR value. During the interpolation process, sensor data reliability weights can be introduced to eliminate outliers and ensure the reliability of the interpolation results.

[0077] In unshaded areas, where there is no cloud cover, direct radiation dominates, and diffuse solar radiation (DH) data shows a significant correlation with diffuse solar radiation (DNR). Therefore, DH data collected by low-density diffuse solar radiation sensors in unshaded areas and baseline monitoring point data with accurately estimated DNR are selected as the basis for extrapolation. The computer equipment can employ either linear or polynomial extrapolation methods. The linear extrapolation method assumes a linear relationship between DH and DNR. Based on the DH-DNR correspondence of the baseline monitoring points, a linear regression model is constructed, and the DNR value is extrapolated from the DH data of the monitoring point to be estimated. The polynomial extrapolation method, for unshaded areas with slightly undulating terrain, constructs a quadratic or cubic polynomial model to fit the nonlinear relationship between DH and DNR, improving extrapolation accuracy. The computer equipment can cross-validate the extrapolation results with the DNR values ​​converted from GH data of nearby high-density total solar radiation sensors, adjusting the extrapolation model parameters accordingly.

[0078] Computer equipment can integrate the interpolation results of the shaded area with the extrapolation results of the unshaded area to obtain the final estimated DNR value of all monitoring points in the solar field, forming a global DNR dataset. Then, the DNR dataset is used to generate a direct radiation intensity distribution map in the form of contour maps, pseudo-color maps, etc. The map clearly marks the color labels of different DNR value intervals, the boundaries between the shaded and unshaded areas, and the location distribution of the heliostat cluster, ensuring that the dynamics of the radiation distribution across the entire field are reflected in real time.

[0079] In one embodiment, a method for monitoring and controlling solar field cloud shading based on multi-sensor data fusion may further include adjusting the working status of heliostats within the shading area. Specifically, this process includes: determining the heliostat cluster within the shading area based on a direct radiation intensity distribution map; adjusting the aiming logic of the heliostat cluster within the shading area in real time, and classifying the heliostat cluster within the shading area into low-priority and high-priority heliostats based on the aiming logic; switching the low-priority heliostats to standby mode and the high-priority heliostats to standby heat storage and reflection mode; and, after the clouds recede, sequentially reactivating the high-priority and low-priority heliostats based on the direct solar radiation intensity measurement results.

[0080] The computer equipment can adjust the aiming logic of the heliostat cluster in real time based on the direct radiation intensity distribution map, switching heliostats in the obscured area to standby or reflecting radiation to the backup thermal storage unit. Specifically, based on the DNR distribution map, the heliostat cluster within the obscured area is identified; the heliostats within the obscured area are switched to standby or reflected to the backup thermal storage unit; and once the clouds have moved away from the area, the heliostats corresponding to excess energy are removed based on real-time DNR measurements.

[0081] Based on the direct radiation intensity distribution map, the computer equipment dynamically adjusts the operating status of heliostats within the shaded area, combined with the operating parameters and energy utilization requirements of the heliostat cluster, to effectively avoid energy loss due to cloud shading. Specifically, the computer equipment can extract the specific range of the shaded area, the included heliostat cluster numbers, and the corresponding DNR values ​​from the direct radiation intensity distribution map, identifying the heliostats that need adjustment; it can also obtain the real-time load status of the main receiver and the available capacity of the backup thermal storage unit to determine the feasibility of energy recovery for the heliostats within the shaded area.

[0082] For heliostats with extremely low DNR values ​​and long periods of shading, or those with low reflection efficiency and far from the backup thermal storage unit, the computer equipment issues a standby command to drive the heliostat's azimuth and elevation angle actuators to reset to the preset standby position, shutting down the tracking drive system and retaining only the status monitoring function to reduce ineffective energy consumption. For heliostats with medium DNR values ​​and short periods of shading, or those with high reflection efficiency and close to the backup thermal storage unit, a turning command is issued. Based on the real-time coordinates of the backup thermal storage unit, the target reflection angle is calculated, and the heliostat is driven to adjust its azimuth and elevation angles so that the reflected light is accurately focused onto the heat-absorbing surface of the backup thermal storage unit, realizing the energy recovery of scattered light or remaining direct light. After the heliostat executes the adjustment command, it feeds back the actual working status data to the computer equipment in real time through its onboard position and angle sensors to ensure accurate execution of the adjustment command.

[0083] The computer equipment can continuously monitor the direct radiation intensity distribution map. When the DNR value of the shaded area is higher than the set threshold twice in a row, and the cloud shading trajectory predicts that there will be no new cloud cover in the area, it is determined that the cloud has left. The computer equipment issues a callback command in the order of priority heliostats with high reflectivity, followed by heliostats with low reflectivity, i.e., high priority heliostats and low priority heliostats. The command drives the heliostats to adjust to the target angle of the main receiver in sequence, restarts the tracking drive system, and restores the solar tracking function. During the callback process, the callback rate is dynamically adjusted according to the real-time load of the main receiver to avoid energy overload.

[0084] Step 210: Introduce geographic information system data, calculate and correct the projection length and shape of cloud shadows on the terrain based on cloud shading trajectory and terrain undulation; integrate meteorological satellite data and cloud height data to construct a three-dimensional cloud model, and analyze the impact of cloud height changes on direct solar radiation intensity based on the three-dimensional cloud model to achieve cloud shading monitoring and control.

[0085] Computer equipment can combine geographic information system data to obtain topographic information of solar energy fields, correct shadow lengths under complex terrain, and improve the accuracy of direct radiation estimation under complex terrain.

[0086] In one embodiment, a method for monitoring and controlling cloud shading of a solar field based on multi-sensor data fusion may further include a projection correction process. The specific process includes: introducing geographic information system data to obtain terrain information of the solar field; determining the terrain undulation based on the cloud shading trajectory and the terrain information; and correcting the projection length and shape of the cloud shadow on the terrain using a terrain correction algorithm. The terrain correction algorithm is a ray tracing algorithm or a light tracing algorithm.

[0087] To address the issue of cloud shadow projection estimation bias under complex terrain conditions in solar fields, Geographic Information System (GIS) data can be introduced and combined with cloud shading trajectories to accurately calculate and correct the length and shape of cloud shadow projections on the terrain, ensuring the accuracy of the division between shaded and unshaded areas and the estimation of direct radiation intensity.

[0088] Specifically, computer equipment can use geographic information system software such as ArcGIS and QGIS to acquire three-dimensional terrain data of the entire solar field and a certain surrounding area. Data types include digital elevation models (DEMs), terrain slope, and aspect data to ensure accurate reflection of terrain undulations. Next, the computer equipment can perform format conversion, noise removal, and coordinate calibration on the acquired GIS terrain data to ensure that the coordinate system of the terrain data is consistent with the coordinate system of the solar field monitoring points and sensor network, thus achieving spatial correlation between the terrain data and other monitoring data.

[0089] Next, the computer equipment can calculate the theoretical projection length of the cloud shadow on the horizontal ground based on the cloud shadow trajectory predicted above, combined with the real-time position parameters of the sun, and according to the cloud height data and the sun's altitude angle; it can also combine the boundary coordinates of the cloud shadow trajectory and the sun's azimuth angle to outline the initial projection shape of the cloud shadow on the horizontal ground, forming a two-dimensional initial projection area. Next, considering the topographic undulations of the solar field, the influence of terrain slope and aspect on cloud shadow projection is analyzed. Specifically, slope changes the actual shadow coverage length, and aspect affects the shadow coverage direction. Depressions or convex terrain may lead to localized shadow overlap or absence. Using ray tracing or ray-tracing algorithms, the initial projection area is overlaid with 3D terrain data for analysis, calculating the impact of terrain undulations on shadow projection point by point: For convex terrain, the actual shadow projection length is shortened based on the terrain elevation difference, correcting the shadow boundary to shift away from the sun; for concave terrain, the actual shadow projection length is extended, correcting the shadow boundary to expand towards the sun; for sloping terrain, the shadow projection stretching ratio is adjusted based on the slope angle to ensure the shadow shape matches the terrain contour. Corrected cloud shadow projection length and shape data are generated, and the spatial distribution maps of shaded and unshaded areas are updated to better reflect the actual cloud shading conditions under the terrain.

[0090] In one embodiment, the computer device can also integrate meteorological satellite data to construct a three-dimensional cloud model and analyze the impact of cloud height changes on direct radiation. Specifically, it can integrate meteorological satellite cloud image data to construct a three-dimensional cloud model; combine time series analysis to predict the impact of cloud height changes on direct radiation (DNR); and in multi-layer cloud scenarios, optimize DNR estimation through height stratification.

[0091] In one embodiment, a solar field cloud shading monitoring and control method based on multi-sensor data fusion may further include a periodic calibration process for the sensor network. The specific process includes: periodically calibrating the deployed low-density scattered solar radiation sensor network and high-density overall solar radiation sensor network on-site using a UAV equipped with a standard radiometer to obtain calibration results; and adjusting the parameters of each sensor according to the calibration results.

[0092] Computer equipment can utilize drones equipped with standard radiometers to perform periodic on-site calibration of sensor networks, reducing system construction and maintenance costs. Specifically, a periodic automatic calibration process can be established, such as monthly; drones equipped with standard radiometers can be used to perform on-site calibration of distributed nodes; and the parameters of the sensor network can be adjusted based on the calibration results to reduce system errors.

[0093] In one embodiment, the computer device can also determine whether it is necessary to continue monitoring the solar field cloud shading. If so, it can re-collect scattered solar radiation data and total solar radiation data, estimate the direct solar radiation intensity, predict the cloud shading trajectory, adjust the working status of the heliostat in the shading area, and form a closed-loop monitoring and control. If it is not necessary to continue monitoring the solar field cloud shading, it can end the monitoring process.

[0094] It should be understood that although the steps in the flowchart above are shown sequentially as indicated by the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless explicitly stated herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least some steps in the flowchart above may include multiple sub-steps or multiple stages. These sub-steps or stages are not necessarily completed at the same time, but can be executed at different times. The execution order of these sub-steps or stages is not necessarily sequential, but can be performed alternately or in turn with other steps or at least some of the sub-steps or stages of other steps.

[0095] In one embodiment, such as Figure 3 As shown, a solar field cloud shading monitoring and control system based on multi-sensor data fusion is provided, including: a data acquisition module 310, a direct solar radiation intensity estimation module 320, a cloud shading trajectory prediction module 330, a heliostat adjustment module 340, and an analysis and monitoring module 350, wherein:

[0096] The data acquisition module 310 is used to deploy a low-density scattered solar radiation sensor network and a high-density overall solar radiation sensor network inside the solar field to acquire scattered solar radiation data and overall solar radiation data in different areas of the solar field.

[0097] The direct solar radiation intensity estimation module 320 is used to process the scattered solar radiation data and the total solar radiation data using adaptive filtering technology, and to estimate the direct solar radiation intensity based on the processed solar radiation data through the radiation balance equation.

[0098] The cloud obscuration trajectory prediction module 330 is used to acquire ground and sky image data and cloud height data through fisheye camera and lidar, and to analyze the ground and sky image data and cloud height data using convolutional neural network to predict cloud obscuration trajectory.

[0099] The heliostat adjustment module 340 is used to divide the monitoring point into a shaded area and an unshaded area according to the cloud shading trajectory. It uses the overall solar radiation interpolation and the scattered solar radiation extrapolation method to perform spatial interpolation on the direct solar radiation intensity to obtain a direct radiation intensity distribution map. Based on the direct radiation intensity distribution map, it adjusts the working status of the heliostat in the shaded area.

[0100] The analysis and monitoring module 350 is used to import data from the geographic information system, calculate and correct the projection length and shape of the cloud shadow on the terrain based on the cloud shading trajectory and the terrain undulation; integrate meteorological satellite data and cloud height data to construct a three-dimensional cloud model, and analyze the impact of cloud height changes on the intensity of direct solar radiation based on the three-dimensional cloud model, so as to realize cloud shading monitoring and control.

[0101] In one embodiment, the data acquisition module 310 is further configured to determine the number and location of low-density scattered solar radiation sensors and high-density total solar radiation sensors based on the size of the solar field; install a low-density scattered solar radiation sensor network based on the number and location of the sensors to measure scattered solar radiation data in different areas of the solar field; and install a high-density total solar radiation sensor network based on the number and location of the sensors to measure total solar radiation data in different areas of the solar field.

[0102] In one embodiment, the direct solar radiation intensity estimation module 320 is further configured to estimate the direct solar radiation intensity based on the scattered solar radiation data and the total solar radiation data using the radiation balance equation; to filter sensor noise data using adaptive filtering technology and to introduce a real-time data calibration algorithm to dynamically optimize the sensor data weights, thereby completing the processing of the scattered solar radiation data and the total solar radiation data; wherein, the adaptive filtering technology is selected from one of the Kalman filtering algorithm, the wavelet transform denoising algorithm, and the median filtering algorithm; and the real-time data calibration algorithm is selected from one of the least squares method, Bayesian estimation, and the maximum entropy algorithm.

[0103] In one embodiment, the cloud occlusion trajectory prediction module 330 is further configured to analyze ground and sky image data and cloud height data using a convolutional neural network to extract cloud features; determine cloud boundaries and cloud shadow projections based on cloud height data according to cloud features; predict the speed, direction, and occlusion status of cloud movement based on time series analysis; and predict the cloud occlusion trajectory based on cloud boundaries, cloud shadow projections, and the speed, direction, and occlusion status of cloud movement.

[0104] In one embodiment, the spatial interpolation difference method is selected from inverse distance weighted interpolation and Kriging interpolation; the extrapolation method is selected from linear extrapolation and polynomial extrapolation.

[0105] In one embodiment, the heliostat adjustment module 340 is further configured to determine the heliostat cluster within the shielded area based on the direct solar radiation intensity distribution map; adjust the aiming logic of the heliostat cluster within the shielded area in real time, and classify the heliostat cluster within the shielded area into low-priority heliostats and high-priority heliostats based on the aiming logic; switch the low-priority heliostats to standby mode and switch the high-priority heliostats to standby heat storage and reflection mode; and when the clouds recede, sequentially revert the high-priority heliostats and low-priority heliostats based on the direct solar radiation intensity measurement results.

[0106] In one embodiment, the analysis and monitoring module 350 is further configured to import geographic information system data to obtain terrain information of the solar field; determine the terrain undulation based on the cloud shading trajectory and the terrain information, and correct the projection length and shape of the cloud shadow on the terrain through a terrain correction algorithm; wherein the terrain correction algorithm is a ray tracing algorithm or a light tracing algorithm.

[0107] In one embodiment, a sensor calibration module is also included, which is used to periodically calibrate the deployed low-density scattered solar radiation sensor network and high-density overall solar radiation sensor network on-site using a standard radiometer mounted on a UAV, and obtain calibration results; and adjust the parameters of each sensor according to the calibration results.

[0108] In one embodiment, a cyclic monitoring module is also included to determine whether it is necessary to continue monitoring the solar field cloud shading. If so, the scattered solar radiation data and total solar radiation data are collected again, the direct solar radiation intensity is estimated, the cloud shading trajectory is predicted, and the working status of the heliostat in the shading area is adjusted to form a closed-loop monitoring and control. If it is not necessary to continue monitoring the solar field cloud shading, the monitoring is terminated.

[0109] In one embodiment, a computer device is provided, which may be a terminal, and its internal structure diagram may be as follows: Figure 4As shown, the computer device includes a processor, memory, network interface, display screen, and input devices connected via a system bus. The processor provides computing and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system and computer programs. The internal memory provides an environment for the operation of the operating system and computer programs stored in the non-volatile storage media. The network interface is used to communicate with external terminals via a network connection. When executed by the processor, the computer program implements a method for monitoring and controlling solar field cloud shading based on multi-sensor data fusion. The display screen can be an LCD screen or an e-ink display screen. The input devices can be a touch layer covering the display screen, buttons, a trackball, or a touchpad on the computer device's casing, or an external keyboard, touchpad, or mouse.

[0110] Those skilled in the art will understand that Figure 4 The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the computer device to which the present application is applied. Specific computer devices may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0111] In one embodiment, a computer device is provided, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the steps of a method for monitoring and controlling solar field cloud shading based on multi-sensor data fusion.

[0112] In one embodiment, a computer-readable storage medium is provided having a computer program stored thereon, which, when executed by a processor, implements the steps of a method for monitoring and controlling solar field cloud shading based on multi-sensor data fusion.

[0113] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium. When executed, the computer program can include the processes of the embodiments of the above methods. Any references to memory, storage, databases, or other media used in the embodiments provided in this application can include non-volatile and / or volatile memory. Non-volatile memory can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory can include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in various forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), dual data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), Rambus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), etc.

[0114] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0115] The embodiments described above are merely illustrative of several implementation methods of this application, and while the descriptions are relatively specific and detailed, they should not be construed as limiting the scope of the invention patent. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these all fall within the protection scope of this application. Therefore, the protection scope of this patent application should be determined by the appended claims.

Claims

1. A method for monitoring and controlling cloud shading in solar fields based on multi-sensor data fusion, characterized in that, The method includes: A low-density scattered solar radiation sensor network and a high-density overall solar radiation sensor network are deployed inside the solar field to acquire scattered solar radiation data and overall solar radiation data in different regions of the solar field. Adaptive filtering technology is used to process the scattered solar radiation data and the total solar radiation data. Based on the processed solar radiation data, the direct solar radiation intensity is estimated using the radiation balance equation. Ground and sky image data and cloud height data are acquired by fisheye camera and lidar, and convolutional neural network is used to analyze the ground and sky image data and cloud height data to predict cloud obscuring trajectory. Based on the cloud obscuration trajectory, the monitoring points are divided into obscured and unobscured areas. The direct solar radiation intensity is spatially interpolated using the overall solar radiation interpolation and the scattered solar radiation extrapolation methods to obtain a direct radiation intensity distribution map. Based on the direct radiation intensity distribution map, the working status of the heliostats in the obscured area is adjusted. By introducing geographic information system data, the projected length and shape of the cloud shadow on the terrain are calculated and corrected based on the cloud shading trajectory and terrain undulation. By fusing meteorological satellite data and cloud height data, a three-dimensional cloud model is constructed, and the impact of cloud height changes on the intensity of direct solar radiation is analyzed based on the three-dimensional cloud model, so as to realize cloud shading monitoring and control.

2. The solar field cloud shading monitoring and control method based on multi-sensor data fusion according to claim 1, characterized in that, A low-density scattered solar radiation sensor network and a high-density overall solar radiation sensor network are deployed within the solar field to acquire scattered solar radiation data and overall solar radiation data for different regions of the solar field, including: Based on the size of the solar field, determine the number and location of low-density scattered solar radiation sensors and high-density overall solar radiation sensors. Based on the number and location of the aforementioned sensors, a low-density scattered solar radiation sensor network is installed to measure scattered solar radiation data in different areas of the solar field. Based on the number and location of the sensors, a high-density total solar radiation sensor network is installed to measure total solar radiation data in different areas of the solar field.

3. The solar field cloud shading monitoring and control method based on multi-sensor data fusion according to claim 1, characterized in that, Adaptive filtering techniques are used to process the scattered solar radiation data and the total solar radiation data. Based on the processed solar radiation data, the direct solar radiation intensity is estimated using the radiation balance equation, including: Based on the scattered solar radiation data and the total solar radiation data, the intensity of direct solar radiation is estimated using the radiation balance equation. Adaptive filtering technology is used to filter sensor noise data, and a real-time data calibration algorithm is introduced to dynamically optimize sensor data weights, thereby completing the processing of the scattered solar radiation data and the total solar radiation data. The adaptive filtering technique is selected from one of the following: Kalman filtering algorithm, wavelet transform denoising algorithm, and median filtering algorithm; the real-time data calibration algorithm is selected from one of the following: least squares method, Bayesian estimation, and maximum entropy algorithm.

4. The solar field cloud shading monitoring and control method based on multi-sensor data fusion according to claim 1, characterized in that, Analyzing the ground and sky image data and cloud height data using a convolutional neural network to predict cloud obscuring trajectories includes: The ground and sky image data and cloud height data are analyzed using a convolutional neural network to extract cloud features; Based on the cloud characteristics, cloud boundaries and cloud shadow projections are determined using the cloud height data. Based on time series analysis, predict the speed, direction, and occlusion of cloud movement; Based on the cloud boundary, cloud shadow projection, cloud movement speed, direction, and occlusion status, the cloud occlusion trajectory is predicted.

5. The solar field cloud shading monitoring and control method based on multi-sensor data fusion according to claim 1, characterized in that, The spatial interpolation difference method is selected from one of inverse distance weighted interpolation and Kriging interpolation; the extrapolation method is selected from one of linear extrapolation and polynomial extrapolation.

6. The solar field cloud shading monitoring and control method based on multi-sensor data fusion according to claim 1, characterized in that, Based on the direct radiation intensity distribution map, adjust the operating status of the heliostats within the shielded area, including: The cluster of heliostats within the shielded area is determined based on the direct radiation intensity distribution map. The aiming logic of the heliostat cluster within the shielded area is adjusted in real time, and the heliostat cluster within the shielded area is divided into low-priority heliostats and high-priority heliostats based on the aiming logic. Switch the low-priority heliostat to standby mode and switch the high-priority heliostat to standby heat storage and reflection mode. Once the clouds have dispersed, the high-priority heliostats and low-priority heliostats will be switched back to their original positions based on the direct solar radiation intensity measurement results.

7. The method for monitoring and controlling cloud shading of solar fields based on multi-sensor data fusion according to claim 1, characterized in that, Introducing geographic information system data, based on the cloud shading trajectory, the projected length and shape of the cloud shadow on the terrain are calculated and corrected according to the terrain undulation, including: Geographic information system data is introduced to obtain the topographic information of the solar field; Based on the cloud occlusion trajectory, the terrain undulation is determined according to the terrain information, and the length and shape of the cloud shadow projection on the terrain are corrected by the terrain correction algorithm. The terrain correction algorithm is either a ray tracing algorithm or a ray tracing algorithm.

8. The method for monitoring and controlling cloud shading of solar fields based on multi-sensor data fusion according to claim 1, characterized in that, The method further includes: The deployment of low-density scattered solar radiation sensor network and high-density overall solar radiation sensor network was periodically calibrated on-site using a UAV equipped with a standard radiometer, and the calibration results were obtained. Adjust the parameters of each sensor based on the calibration results.

9. The method for monitoring and controlling cloud shading of solar fields based on multi-sensor data fusion according to claim 1, characterized in that, The method further includes: Determine whether it is necessary to continue monitoring the solar field cloud shading. If so, collect scattered solar radiation data and total solar radiation data again, estimate the direct solar radiation intensity, predict the cloud shading trajectory, adjust the working status of the heliostat in the shading area, and form a closed-loop monitoring and control system. If continued monitoring of solar field cloud shading is no longer required, then the process ends.

10. A solar field cloud shading monitoring and control system based on multi-sensor data fusion, characterized in that, The system includes: The data acquisition module is used to deploy a low-density scattered solar radiation sensor network and a high-density overall solar radiation sensor network inside the solar field to acquire scattered solar radiation data and overall solar radiation data in different areas of the solar field. The direct solar radiation intensity estimation module is used to process the scattered solar radiation data and the total solar radiation data using adaptive filtering technology, and to estimate the direct solar radiation intensity based on the processed solar radiation data through the radiation balance equation. The cloud obscuration trajectory prediction module is used to acquire ground and sky image data and cloud height data through fisheye camera and lidar, and to analyze the ground and sky image data and cloud height data using convolutional neural network to predict cloud obscuration trajectory. The heliostat adjustment module is used to divide the monitoring point into a shaded area and an unshaded area according to the cloud shading trajectory, and to perform spatial interpolation of the direct solar radiation intensity using the overall solar radiation interpolation and the scattered solar radiation extrapolation method respectively to obtain a direct radiation intensity distribution map. Based on the direct radiation intensity distribution map, the module adjusts the working state of the heliostat in the shaded area. The analysis and monitoring module is used to import data from the geographic information system, calculate and correct the projection length and shape of the cloud shadow on the terrain based on the cloud shading trajectory and the terrain undulation; integrate meteorological satellite data and cloud height data to construct a three-dimensional cloud model, and analyze the impact of cloud height changes on the intensity of direct solar radiation based on the three-dimensional cloud model, so as to realize cloud shading monitoring and control.

Images