Photovoltaic electric field sand and dust meteorological gradient three-dimensional monitoring and early warning system

Abstract

The invention discloses a photovoltaic electric field sand and dust meteorological gradient three-dimensional monitoring and early warning system. According to the system, a plurality of gradient micro-meteorological station networks are arranged in a peripheral area, an edge area and a central area of a photovoltaic electric field, wind speed, wind direction, air temperature, humidity, air pressure, precipitation and radiation are measured at different heights, and soil temperature, humidity and conductivity sensors are buried; piezoelectric wind erosion sensors with different heights from the ground are arranged on the ground surface, and the sensors have particle counting, kinetic energy and single particle energy output and are provided with adjustable gains. The system is also equipped with a multi-rotor unmanned aerial vehicle to measure meteorological and dust concentration data at different heights above the photovoltaic panel. All data are recorded in a unified mode through the collecting unit and processed through a fusion algorithm, the intelligent early warning processing unit determines and calculates the dust damage degree through machine learning dynamic weight according to an improved damage index, and early warning is triggered when the index exceeds a threshold value. According to the invention, three-dimensional monitoring and intelligent early warning of sand and dust activities of the photovoltaic electric field can be provided.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of environmental monitoring and disaster warning of solar photovoltaic power stations, and in particular to a photovoltaic power plant sand weather gradient three-dimensional monitoring and warning system. BACKGROUND

[0002] With the construction of large-scale desert photovoltaic bases, wind and sand erosion has become an important factor restricting the service life and power generation efficiency of photovoltaic modules. Sand particle deposition can reduce the solar radiation transmittance, and friction and impact can damage the surface of the module and even cause short circuit. The existing photovoltaic field environmental monitoring mainly focuses on the contamination or power generation change of the module surface, and the cleaning demand is inferred by comparing the output of the standard module. Such a scheme does not directly monitor the sand transport process, and cannot reveal the wind field difference inside and outside the photovoltaic panel and the sand lifting threshold, so it is difficult to issue a sand disaster warning in a timely manner.

[0003] In the field of wind and sand observation, common monitoring facilities include a single-height meteorological tower or an integrated wind and sand monitoring station, usually equipped with wind speed sensors, temperature and humidity sensors, and automatic sand collectors. However, single-tower or single-layer monitoring cannot reflect the wind field gradient and sand distribution in different areas and different heights of the photovoltaic field, and is unable to capture the sand lifting critical wind speed and sand transport mechanism. In addition, although some studies use multi-rotor unmanned aerial vehicles to carry sensors to obtain atmospheric parameters at different heights, they often do not analyze with ground stations, and lack special evaluation models for photovoltaic fields.

[0004] In summary, there is an urgent need for a three-dimensional monitoring and intelligent warning system for photovoltaic power plants that can simultaneously obtain ground and air multi-element information and accurately evaluate wind and sand hazards through cross-domain data fusion and dynamic threshold algorithms. SUMMARY

[0005] Technical purpose: In view of the problems that the existing photovoltaic power plant sand activity cannot be monitored in multiple layers in real time, the sand lifting threshold is difficult to determine, and the hazard evaluation lacks adaptability, the present application discloses a photovoltaic power plant sand weather gradient three-dimensional monitoring and warning system, which can comprehensively master the wind and sand dynamics of different areas of the photovoltaic field and timely issue warnings by constructing a layered monitoring station network, introducing a piezoelectric wind erosion sensor with single particle energy output and adjustable gain, and using unmanned aerial vehicle three-dimensional detection combined with machine learning algorithm to dynamically calculate the hazard index.

[0006] Technical scheme: In order to achieve the above technical purpose, the present application adopts the following technical scheme:

[0007] A photovoltaic power plant sand weather gradient three-dimensional monitoring and warning system, comprising:

[0008] A gradient micro-meteorological station network is arranged in the peripheral area, edge area and central area of the photovoltaic power station, each monitoring device includes a support tower, an air temperature sensor, a relative humidity sensor and a rain gauge are arranged at the near-surface layer measuring height, an atmospheric pressure sensor, a four-component radiation sensor and a net radiation sensor are arranged at the platform layer measuring height, a wind speed sensor and a wind direction sensor are arranged at the near-surface layer measuring height and the free flow layer measuring height respectively to form a vertical wind field gradient of the near-surface layer under the photovoltaic panel and the free flow layer above the photovoltaic panel, and soil temperature, soil volume moisture content and soil conductivity sensors are arranged on the ground surface and at different depths underground;

[0009] A piezoelectric wind erosion sensing unit is arranged near the ground surface of each monitoring device, at least two piezoelectric wind erosion sensors with different heights from the ground are arranged, each piezoelectric wind erosion sensor has particle count pulse output, kinetic energy pulse output and pulse amplitude output, and is provided with adjustable gain;

[0010] An aerial mobile detection unit includes at least one multi-rotor unmanned aerial vehicle, which carries sensors for measuring air temperature, relative humidity, wind speed, wind direction, air pressure and dust concentration, and can move to measure at different heights above the photovoltaic panel;

[0011] A data acquisition and processing unit is used to receive and store raw data from the gradient micro-meteorological station network, the piezoelectric wind erosion sensing unit and the aerial mobile detection unit, and can remotely control the gain of the piezoelectric wind erosion sensor;

[0012] An intelligent early warning processing unit is used to fuse the raw data, calculate the dust hazard index based on the hazard index formula, and output the dust early warning signal when the hazard index reaches the threshold value.

[0013] Preferably, the piezoelectric wind erosion sensor is a piezoelectric wafer sensor, which provides three outputs of particle count pulse, kinetic energy pulse and pulse amplitude, and has two gain settings of 1× and 10×; the wafer extends to the edge of the shell to form a sealed structure, so that the radial response is uniform and multiple counting is not generated.

[0014] Preferably, the data acquisition and processing unit receives raw data at a level of seconds to hours through wired or wireless means, and can remotely control the gain of the piezoelectric wind erosion sensor, so that it uses 10× gain at low dust intensity and 1× gain at high dust intensity.

[0015] Preferably, the intelligent early warning processing unit includes a data fusion module for time synchronization, quality control, unit unification and spatial weight fusion of raw data from different sources; a hazard index calculation module for calculating the dust hazard index according to the wind field disturbance term, the concentration risk term, the wind erosion intensity term and the soil moisture term; a threshold adaptive module for determining the hazard index trigger threshold according to historical statistics or cost function and online correction based on soil moisture content and friction velocity working conditions.

[0016] Preferably, the hazard index calculation formula is:

[0017] ,

[0018] wherein is the average wind speed outside the photovoltaic panel, is the average wind speed under the photovoltaic panel, is the average sand concentration in the monitoring area, is the sand concentration threshold value, is the kinetic energy flux per unit time, is the reference kinetic energy flux threshold value, is the soil volume water content or humidity index, is the soil humidity reference value, is the weight coefficient trained according to historical monitoring data; when , the intelligent early warning processing unit outputs a sandstorm early warning signal, wherein is the hazard index triggering threshold value.

[0019] Preferably, the calculation formula of the kinetic energy flux is:

[0020] ,

[0021] wherein is the kinetic energy flux per unit time, is the kinetic energy pulse count of the piezoelectric wind erosion sensor, is the single pulse energy, is the measurement time.

[0022] Preferably, the intelligent early warning processing unit fuses and denoises the multi-source data through Kalman filtering, polynomial regression or machine learning algorithm, and dynamically adjusts the hazard index triggering threshold value according to the historical training model of the weight coefficient.

[0023] Preferably, the intelligent early warning processing unit identifies the sand-raising wind speed according to the change trend of the piezoelectric wind erosion sensor particle count and the wind speed data, and calculates the shear wind speed by using the shear wind speed formula to dynamically update the reference kinetic energy flux threshold value and the hazard index triggering threshold value , wherein is the shear wind speed, and are the wind speeds at two measuring heights, and are the corresponding installation heights of the piezoelectric wind erosion sensors, is the von Karman constant.

[0024] Preferably, the aerial mobile detection unit cruises at at least three different altitudes above the photovoltaic panel along a preset route, and its sensors include particle size measurement or laser scattering dust meters to generate a three-dimensional distribution of dust concentration and fuse it with ground concentration data.

[0025] A three-dimensional monitoring and early warning method for dust meteorological gradients in photovoltaic power fields, applied to the aforementioned three-dimensional monitoring and early warning system for dust meteorological gradients in photovoltaic power fields, specifically includes the following steps:

[0026] Gradient micrometeorological monitoring devices are deployed in different areas of the photovoltaic power plant to collect data on wind speed, wind direction, temperature and humidity, air pressure, precipitation, radiation, soil temperature and humidity, and electrical conductivity.

[0027] At least two piezoelectric wind erosion sensors at different heights are installed near the ground surface of each monitoring device to collect data on the particle count, kinetic energy pulse, and pulse amplitude of sand impacts.

[0028] Multi-rotor drones were used to cruise at different altitudes above the photovoltaic panels to measure air temperature, humidity, wind speed, wind direction, air pressure, and dust concentration.

[0029] All sensor data is collected and stored through the data acquisition and processing unit, and the gain of the piezoelectric wind erosion sensor is remotely controlled.

[0030] In the intelligent early warning processing unit, the collected data is fused and denoised, and the hazard index I is calculated based on weighting coefficients. Dust storm warnings were issued periodically;

[0031] By analyzing the inflection point of the particle count versus wind speed curve of a piezoelectric wind erosion sensor, the sand-raising wind speed is identified, and the reference kinetic energy flux threshold is updated by combining it with shear wind speed calculation. and hazard threshold And the sensor sensitivity is adjusted via remote gain control.

[0032] Beneficial Effects: The photovoltaic electric field dust meteorological gradient three-dimensional monitoring and early warning system provided by this invention has the following beneficial effects:

[0033] 1. This invention achieves three-dimensional coverage by deploying three sets of gradient micro-meteorological stations in the outer / edge / central areas and linking them with a mobile aerial detection unit. Through unified time synchronization, quality control, and unit standardization by the data acquisition and processing unit, phase and scale differences between multi-source devices are eliminated. Compared with existing single-point / single-layer observations, this invention utilizes a dual-gradient comparison of "outer → inner" and "lower → upper" gradients to distinguish the contributions of external transport and internal dust generation, capturing the differences in dust generation time, transport path, and inter- and sub-plate ventilation disturbance mechanisms. Thus, without increasing sensor accuracy requirements, spatial attribution and identification of dust sources are achieved through geometric deployment and data organization, providing interpretable upwind / downwind evidence chains for subsequent early warning.

[0034] 2. This invention introduces a piezoelectric wind erosion sensing unit and integrates it with soil temperature / volume moisture content / conductivity profiles. It calculates wind erosion kinetic energy flux and particle activity intensity based on impact pulse energy conversion, rather than solely relying on near-ground wind speed or empirical thresholds. The hazard index calculation module then incorporates wind speeds from two layers and surface / aerial concentration fields into the calculation. Compared to existing methods that "fix the wind speed threshold for sand-blowing," this invention's quantification method simultaneously reflects incoming kinetic energy and surface moisture / cementation status. When the surface is moist or crusted, it can suppress false alarms; when the surface is dry and loose with enhanced under-slab ventilation, it can reflect the activation of fine particles in advance. Therefore, it avoids the two-way hysteresis of wind speed reaching the threshold without sand-blowing and sand-blowing without wind speed reaching the threshold, significantly improving the response sensitivity and reliability of real sand-blowing events.

[0035] 3. This invention establishes a four-component interpretable structure for the dust hazard index I. Each component is provided with consistent time-space point values ​​by the data fusion module, which also outputs the component contribution. In the threshold adaptive module, quantile statistics and cost-optimal criteria are adopted, and state variables such as shear wind speed and soil moisture content are explicitly introduced to perform online correction and anti-shaking processing on the hazard index trigger threshold. This ensures that under conditions such as seasonal changes in incoming flow, seasonal fluctuations in surface moisture content, and changes in the underlying surface of the array, the warning triggering conditions adapt to the changing operating conditions, keeping false alarms and missed alarms under control. Accordingly, the output dust warning signal (graded / linked / recorded) has the comprehensive advantages of stable thresholds, interpretable sources, and implementable strategies. It is easy to link with site operation and maintenance (such as under-floor cleaning, boundary protection, temporary restricted operations, etc.) to achieve earlier triggering and clearer handling guidance under the same false alarm constraints. Attached Figure Description

[0036] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the accompanying drawings used in the description of the embodiments or the prior art will be briefly introduced below.

[0037] Figure 1 This is a block diagram of the overall system structure of the present invention;

[0038] Figure 2 This is a flowchart of the method of the present invention;

[0039] Figure 3 A schematic diagram of the flight path of an aerial mobile detection unit (UAV).

[0040] Figure 4 This is a schematic diagram of a three-dimensional dust concentration point cloud and slice. Detailed Implementation

[0041] The present invention will now be described more clearly and completely by way of a preferred embodiment in conjunction with the accompanying drawings, but this does not limit the invention to the scope of the described embodiment.

[0042] Near-surface elevation measurement refers to the measurement height range close to the ground and directly affected by surface energy and roughness. It mainly characterizes the thermal and turbulent exchange characteristics of the near-surface layer under or between photovoltaic panels. In this invention, this layer is used to deploy elements and equipment sensitive to surface effects, such as air temperature and relative humidity sensors (equipped with ventilated radiation shields) and rain gauges, as well as a layer of wind speed and direction sensors under / between the panels. The near-surface elevation measurement should be located in a position that is accessible to personnel and not easily affected by ground re-radiation, vegetation splashing, or backflow of accumulated water. It should also meet the following requirements: maintain sufficient horizontal clearance from the ground, photovoltaic supports, and obstacles; and ensure that the area around the sensor is as open and well-ventilated as possible to prevent local backflow and dead zones from affecting the readings. Based on conventional meteorological deployment experience, temperature and humidity sensors are usually installed in the range of approximately 1.2-1.8m, and rain gauges are usually installed in the range of approximately 0.6-0.8m, requiring them to be horizontal, stable, and equipped with windproof / splashproof accessories. These ranges are only for guiding the implementation example.

[0043] Platform layer height measurement refers to the height layer on the side arm or platform of the monitoring tower used to install instruments that require stable mounting bases and are sensitive to obstruction / albedo. In this invention, atmospheric pressure sensors, four-component radiation and net radiation sensors are preferentially installed on this layer, placing them in a relatively stable, unobstructed position that facilitates calibration and maintenance; simultaneously, it can form a vertical thermal radiation gradient with the near-surface temperature and humidity. Platform layer height measurement should ensure that: the sensor field of view is not obstructed by components, supports, or cables; the horizontality and azimuth of the radiation sensors meet the requirements; and a reasonable clearance is maintained from ground heat sources or reflective surfaces to reduce systematic errors caused by surface albedo and thermal disturbance. In engineering, the platform layer is often arranged on the same layer as the side arm of the tower at approximately 1.5-2.5m (placed vertically above and below the near-surface temperature and humidity layer) to reduce mutual interference and facilitate wiring and protection.

[0044] The free-flow layer (upper layer) elevation measurement refers to the height layer located above the top edge of the photovoltaic module, largely free from the influence of under-panel backflow and wake. It is primarily used to deploy a second layer of wind speed / direction sensors and multiple elevation reference points for drone navigation, forming a vertical gradient comparison with the near-surface wind field. This layer reflects the coupling between external inflow and internal disturbances, and is crucial for determining the critical wind speed for sand initiation, shear wind speed, and the difference in wind field between the module and its underside. The free-flow layer elevation measurement should be a certain clearance above the upper edge of the module, avoiding the wake of the module and the flow around the tower; a typical practice is to be approximately 0.5-2m above the upper edge of the module (determined based on the site array spacing and wind tunnel / on-site calibration), ensuring that the sensors face the wind and are away from cables and components that may obstruct them. During drone navigation, multiple upper-layer heights (e.g., 5m, 15m, 25m, etc.) are selected for multi-layer sampling for three-dimensional concentration reconstruction and elevation correction using threshold adaptive algorithms.

[0045] Near-surface elevation measurement is used to obtain the dominant heat, humidity and precipitation inputs below the plate, between the plate and the surface, as well as a layer of wind field; platform elevation measurement provides reference quantities such as radiation and air pressure for stable installation and unobstructed field of view; free flow layer elevation measurement is used to obtain the free flow above the component and its vertical gradient with the near-surface layer. The three layers can simultaneously identify: (1) the weakening, acceleration or reversal of the incoming flow after entering the field; (2) the backflow, channel effect and sand-raising triggering conditions below / between the plate; (3) the influence of radiation and thermal boundary conditions on sand and dust transport; and serve as component inputs for the hazard index and threshold adaptive module of this invention, improving the accuracy and interpretability of the early warning.

[0046] like Figure 1 As shown, a three-dimensional monitoring and early warning system for meteorological gradients of sand and dust in photovoltaic power plants includes:

[0047] A gradient micro-weather station network is established, deploying multiple monitoring devices in the outer, edge, and central areas of the photovoltaic power plant. Each monitoring device includes a supporting tower and sensor arrangement. Air temperature sensors, relative humidity sensors, and rain gauges are installed at the near-surface elevation level; atmospheric pressure sensors, four-component radiation sensors, and net radiation sensors are installed at the platform elevation level; wind speed and wind direction sensors are installed at at least two elevation levels to establish the vertical gradient between the near-surface layer below the plant and the free-flow layer above it. The two elevation levels refer to the near-surface elevation and the free-flow layer elevation. Rain gauges use tipping bucket or weighing rain sensors, with the center of the opening approximately 0.7m above the ground (or 1.2m in northern regions) to ensure the continuity and comparability of the observation data. Air temperature and relative humidity sensors are typically installed at a height of approximately 1.5m; atmospheric pressure sensors, four-component radiation sensors, and net radiation sensors are installed at a platform elevation of approximately 2m; and wind speed and wind direction sensors are installed at the near-surface layer (e.g., 0.5m) and the free-flow layer (e.g., 8m or 10m), respectively. Soil temperature, volumetric moisture content, and soil electrical conductivity sensors are installed at different depths on the surface and underground (e.g., 0cm, −5cm, −10cm, and −2.5cm) to monitor soil moisture gradients.

[0048] The piezoelectric wind erosion sensing unit employs at least two piezoelectric wind erosion sensors at different ground heights (e.g., 5cm and 10cm) near the ground surface of each monitoring device to record the number and amplitude of sand particle impact pulses in real time. The piezoelectric wind erosion sensors utilize a ring-shaped piezoelectric crystal structure, outputting particle count (PC), particle kinetic energy (KE), and pulse amplitude (PHA) signals. A wide dynamic range measurement is achieved for both weak and strong sand-raising processes via remote gain switching, and selectable 1× and 10× gain functions are provided. Calibration allows the pulse count and particle kinetic energy to be converted into wind erosion kinetic energy flux.

[0049] The aerial mobile detection unit consists of at least one multi-rotor UAV as the aerial detection device, equipped with sensors for air temperature, relative humidity, wind speed, wind direction, air pressure, and particulate matter concentration. The UAV cruises at different altitudes (e.g., 5m, 15m, 25m) above the photovoltaic panels along a preset route, collecting airborne dust concentration and atmospheric parameters, and transmitting the data back in real time. The multi-altitude dust concentration and micro-meteorological data acquired by the UAV are rasterized and fused with the ground station network to generate a three-dimensional point cloud and slice distribution of dust concentration in the photovoltaic field, used to analyze the differences between external transport and internal dust generation.

[0050] The data acquisition and processing unit receives and stores data from all sensors, supporting adjustable sampling steps at the second and minute levels. This unit can remotely control the gain (1× or 10×) of the wind erosion sensor and perform preliminary data correction and time synchronization.

[0051] The intelligent early warning processing unit includes a data fusion module, a hazard index calculation module, and a threshold adaptation module. The data fusion module performs time alignment, quality control, noise reduction, unit unification, and spatial weighting fusion on raw data from wind speed and direction sensors, piezoelectric wind erosion sensors, soil temperature, humidity, and conductivity sensors, four-component radiation and net radiation sensors, and sensors mounted on a UAV, outputting a synchronous feature vector within a preset time window. The hazard index calculation module calculates the comprehensive hazard index I and its component contributions based on the feature vector using a preset mathematical model, and outputs the index change rate. The threshold adaptation module determines the hazard index trigger threshold I under the constraints of the target false alarm rate or cost function. th It also performs online adaptive correction based on soil moisture content, shear wind velocity, and background concentration drift, using I and I th The comparison outputs a graded early warning signal. This unit acquires preprocessed data from the data acquisition and processing unit, performs noise reduction and completion of wind speed, concentration, and soil parameters using multi-sensor fusion algorithms (such as Kalman filtering or machine learning regression), and then calculates the degree of dust hazard based on the improved hazard index formula:

[0052]

[0053] in The average wind speed outside the photovoltaic panel. The average wind speed under the photovoltaic panel, ratio Measuring wind speed gradient and ventilation disturbance To monitor the regional average dust concentration, the data was obtained by fusing ground-based particulate matter concentration sensors and laser scattering dust data collected by drones. The threshold for dust concentration. The kinetic energy flux per unit time is obtained by multiplying the number of kinetic energy pulses by the energy of a single pulse and dividing by the measurement time. For reference kinetic flux threshold, Soil volumetric moisture content, or humidity index, reflects the inhibitory effect of soil moisture content on sand lifting. This is the baseline value for soil moisture. The weighting coefficients are trained based on historical monitoring data and updated with seasonality and environmental changes; when At that time, among them The intelligent early warning processing unit outputs a sandstorm warning signal based on the threshold of the hazard index. The warning level can be divided into light, moderate and heavy levels according to the magnitude of I, and each level is indicated by a different color.

[0054] The threshold adaptation and remote calibration rules of this invention are as follows:

[0055] Sand-inducing threshold update: By analyzing the changes in particle count and wind speed curves of piezoelectric wind erosion sensors, the wind speed corresponding to a significant increase in particle count is identified as the sand-inducing wind speed threshold; combined with the shear wind speed formula... Calculate shear wind speed to dynamically update the reference kinetic flux threshold. and hazard index trigger threshold This allows the early warning model to adapt to changes in soil moisture and wind direction, among which For shear wind speed, and The wind speeds at two different elevations are as follows: and To correspond to the installation height of the piezoelectric wind erosion sensor, is the von Kármán constant.

[0056] Remote gain control: The data acquisition and processing unit automatically switches the gain of the piezoelectric wind erosion sensor to 1× or 10× based on the real-time sandstorm intensity, ensuring measurement sensitivity and accuracy under low energy and strong wind / sand conditions. Gain switching is achieved by the data acquisition and processing unit through control output wiring.

[0057] Sensor Self-Test and Calibration: In this invention, a benchmark wind and sand flux calibration device (such as a general-purpose field wind and sand collector or a gravity sampling device with equivalent function) can be used in the field to perform quality comparison on the kinetic energy output and pulse amplitude output of the piezoelectric wind erosion sensor. By comparing the mass flux captured by the benchmark device with the sensor pulse energy flux conversion value within the same time window, the equivalent kinetic energy E0 per unit pulse and the effective sensing cross section A are calibrated. eff This device is for calibration only and is not an essential component of the system of this invention. The system periodically compares the kinetic energy pulses of the piezoelectric wind erosion sensor with the quality data of the benchmark wind and sand flux calibration device to correct the sensor sensitivity; it also compares the three-dimensional concentration data collected by UAV with the ground concentration data to verify the spatial distribution model.

[0058] like Figure 2 As shown, a three-dimensional monitoring and early warning method for dust meteorological gradients in photovoltaic power fields, applied to the aforementioned three-dimensional monitoring and early warning system for dust meteorological gradients in photovoltaic power fields, specifically includes the following steps:

[0059] Gradient micrometeorological monitoring devices are deployed in different areas of the photovoltaic power plant to collect data on wind speed, wind direction, temperature and humidity, air pressure, precipitation, radiation, soil temperature and humidity, and electrical conductivity.

[0060] At least two piezoelectric wind erosion sensors at different heights are installed near the ground surface of each monitoring device to collect data on the particle count, kinetic energy pulse, and pulse amplitude of sand impacts.

[0061] Multi-rotor drones were used to cruise at different altitudes above the photovoltaic panels to measure air temperature, humidity, wind speed, wind direction, air pressure, and dust concentration.

[0062] All sensor data is collected and stored through the data acquisition and processing unit, and the gain of the piezoelectric wind erosion sensor is remotely controlled.

[0063] In the intelligent early warning processing unit, the collected data is fused and denoised, and the hazard index I is calculated based on weighting coefficients. Dust storm warnings were issued periodically;

[0064] By analyzing the inflection point of the particle count versus wind speed curve of a piezoelectric wind erosion sensor, the sand-raising wind speed is identified, and the reference kinetic energy flux threshold is updated by combining it with shear wind speed calculation. and hazard threshold And the sensor sensitivity is adjusted via remote gain control.

[0065] Example

[0066] Taking a large desert photovoltaic power plant as an example, the site is divided into an outer zone, an edge zone, and a central zone according to the prevailing wind direction. A gradient micro-meteorological monitoring device, including supporting towers and sensor arrangements, is deployed in each zone. According to the design of this invention, the typical installation positions of each sensor are as follows:

[0067] Near-Ground Height Measurement (NGL): This layer is located within 0.3m to 0.8m above the ground. In this embodiment, a rain gauge (tipping bucket or weighing type) is installed at approximately 0.7m; an air temperature sensor and a relative humidity sensor are installed at approximately 1.5m, and a ventilated radiation shield is provided to reduce the effects of radiation and splashing.

[0068] Platform Level Height Measurement (PL): This level is located on the side arm platform of the support tower, typically at a height of approximately 2 meters. This level houses instruments such as atmospheric pressure sensors, four-component radiation sensors, and net radiation sensors. The platform must be level and unobstructed to ensure the accuracy and stability of the measurements.

[0069] Freeflow stratum height measurement (UL): This layer is located in the freeflow zone above the photovoltaic modules, typically at a height of 8m or 10m. Wind speed and direction sensors are installed in this layer. In conjunction with the anemometers at the near-surface height measurement point, a vertical gradient of the wind field from below the module to above it can be established.

[0070] Soil Profile: Soil sensors are deployed near the base of the monitoring device. Soil temperature sensors are buried at depths of 0 cm, -5 cm, and -10 cm; volumetric moisture content and soil conductivity sensors are buried at depths of -2.5 cm, -5 cm, and -10 cm to capture soil temperature and humidity gradients and changes in conductivity.

[0071] Two piezoelectric wind erosion sensors are deployed on the ground surface at wind directions for each monitoring device, with installation heights of 5 cm and 10 cm respectively. The piezoelectric wind erosion sensors output particle count, kinetic flux, and pulse amplitude signals, and maintain measurement sensitivity under both strong and weak sandstorm conditions through remote gain control.

[0072] All sensor signals are connected to the data acquisition and processing unit, which supports variable sampling steps of 1s, 1min, 30min, and 1h. The unit performs time alignment, quality control, and unit standardization on the raw data, and stores the quality-controlled data locally or on a cloud server. Simultaneously, this unit can execute remote gain switching commands on the piezoelectric wind erosion sensor.

[0073] The aerial mobile detection unit employs a multi-rotor drone, with onboard modules including sensors for temperature, humidity, wind speed, wind direction, air pressure, and particulate matter concentration. The drone cruises at different altitudes above the photovoltaic array along a pre-set route. For example, it can fly along the array channel at altitudes of 5m, 15m, and 25m to acquire real-time concentration distribution data. The drone's sampling data is transmitted in real-time via a wireless link to the ground data acquisition and processing unit. After time synchronization with the concentration data from the ground station network, the data is weighted and fused to generate a three-dimensional dust concentration point cloud of the monitoring area. When analyzing a single sandstorm event, the external transport channels and internal sand-generating areas can also be extracted from the three-dimensional concentration data, providing locational basis for early warning.

[0074] like Figure 3 As shown, a multi-rotor drone was used as a mobile detection device. The drone was equipped with sensors for temperature, humidity, wind speed, wind direction, air pressure, particulate matter concentration, and a laser scattering dust meter. Its flight altitude was set at multiple levels, including 5m, 15m, and 25m. The drone cruised above the photovoltaic panels along a preset route, measuring meteorological and dust data at different altitudes in real time, and transmitting the data wirelessly to a ground station. The drone's monitoring time was synchronized with the ground station, and the collected data was used to generate a three-dimensional point cloud distribution of dust concentration. Figure 4 The image shows a schematic diagram of a 3D dust concentration point cloud and slices. The points in the image are concentration sampling points after the aerial mobile detection unit and the ground are merged.

[0075] The steps for data collection and early warning are as follows:

[0076] Step 1: The data acquisition and processing unit acquires raw data from the micro-weather station, wind erosion sensor, and UAV at different time resolutions of 1 second, 1 minute, and 30 minutes. Timestamp synchronization is used, and deviation correction and outlier removal are performed on the sensor data.

[0077] Step 2: The intelligent early warning processing unit fuses the collected data. The data fusion module performs time alignment, quality control, noise reduction, and unit unification on data from near-surface, plateau, and free-spheric wind speed and direction, piezoelectric erosion, rainfall, radiation, soil profiles, and multi-altitude data from UAVs. When fusing multi-altitude concentration data with significant quality differences, near-surface concentrations are weighted first, and high-altitude concentrations acquired by UAVs are used to correct the background transmission.

[0078] Step 3: Count the kinetic energy pulses from the piezoelectric wind erosion sensor. Single pulse energy Given a measurement time T, calculate the kinetic energy flux per unit time:

[0079]

[0080] Step 4: Hazard Index Calculation and Early Warning: Calculate the wind speed gradient ratio Average dust concentration ratio Kinetic flux ratio Compared to soil moisture Substitute these ratios into the hazard index formula I, and then, based on the weights obtained during training... Calculate the composite index. When I reaches or exceeds the threshold... When the time comes, the system will automatically issue a warning and classify the warnings according to the size of the index.

[0081] Step 5: Monitor the inflection point of particle count change with wind speed using the piezoelectric wind erosion sensor in real time to determine the sand-raising wind speed; calculate the shear wind speed using the shear wind speed formula and correct the reference kinetic energy flux threshold. Simultaneously, the concentration threshold is updated based on historical data. and soil moisture baseline The weighting coefficients can also be updated gradually through online learning.

[0082] Step 6: When the sandstorm intensity is weak, the data acquisition and processing unit controls the wind erosion sensor gain to 10× to improve sensitivity; when there is strong sandstorm or a large number of particles impacting, it switches back to 1× to avoid saturation. The system periodically compares the kinetic energy pulse with the quality of the benchmark sandstorm flux calibration device to correct the sensor sensitivity; the three-dimensional concentration measured by the UAV is compared with the ground concentration to verify and correct the concentration model.

[0083] The above description is only a preferred embodiment of the present invention. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the principle of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.

Claims

1. A three-dimensional monitoring and early warning system for meteorological gradients of sand and dust in photovoltaic electric fields, characterized in that, include: The gradient micro-weather station network deploys multiple monitoring devices in the outer, edge, and central areas of the photovoltaic power plant. Each monitoring device includes a supporting tower, and is equipped with air temperature, relative humidity, and rain gauges at the near-surface elevation, atmospheric pressure, four-component radiation, and net radiation sensors at the platform elevation, and wind speed and wind direction sensors at the near-surface and free strata elevations, respectively, to form a vertical wind field gradient between the near-surface layer under the photovoltaic panels and the free strata on the photovoltaic panels. Soil temperature, soil volumetric moisture content, and soil conductivity sensors are installed at different depths on the surface and underground. The piezoelectric wind erosion sensing unit has at least two piezoelectric wind erosion sensors at different heights above the ground near the ground surface of each monitoring device. Each piezoelectric wind erosion sensor has particle counting pulse output, kinetic energy pulse output and pulse amplitude output, and is equipped with adjustable gain. The aerial mobile detection unit consists of at least one multi-rotor drone equipped with sensors for measuring air temperature, relative humidity, wind speed, wind direction, air pressure, and dust concentration, and is capable of moving and measuring at different heights above the photovoltaic panel. The data acquisition and processing unit is used to receive and store raw data from the gradient micro-weather station network, the piezoelectric wind erosion sensing unit and the aerial mobile detection unit, and can remotely control the gain of the piezoelectric wind erosion sensor. The intelligent early warning processing unit is used to fuse and process the raw data, calculate the dust hazard index based on the hazard index formula, and output a dust warning signal when the hazard index reaches the threshold.

2. The photovoltaic power plant dust meteorological gradient three-dimensional monitoring and early warning system according to claim 1, characterized in that, The piezoelectric wind erosion sensor is a piezoelectric wafer sensor that provides three outputs: particle counting pulse, kinetic energy pulse, and pulse amplitude, and has two gain levels: 1× and 10×. Its wafer extends to the edge of the housing to form a sealed structure, which makes the radial response uniform and does not produce multiple counts.

3. The photovoltaic electric field dust meteorological gradient three-dimensional monitoring and early warning system according to claim 2, characterized in that, The data acquisition and processing unit receives raw data from per second to per hour via wired or wireless means, and can remotely control the gain of the piezoelectric wind erosion sensor, so that it uses 10× gain when the sandstorm intensity is low and 1× gain when the sandstorm intensity is high.

4. The photovoltaic power plant dust meteorological gradient three-dimensional monitoring and early warning system according to claim 1, characterized in that, The intelligent early warning processing unit includes: a data fusion module, used to perform time synchronization, quality control, unit unification, and spatial weight fusion on raw data from different sources; a hazard index calculation module, used to calculate the dust hazard index based on wind field disturbance, concentration risk, wind erosion intensity, and soil moisture; and a threshold adaptive module, used to determine the hazard index trigger threshold based on historical statistics or cost functions and to perform online correction based on soil moisture content and friction velocity conditions.

5. A three-dimensional monitoring and early warning system for meteorological gradients of photovoltaic power fields for dust storms according to claim 4, characterized in that, The formula for calculating the hazard index is: ， in The average wind speed outside the photovoltaic panel. The average wind speed under the photovoltaic panel. To monitor the average dust concentration in the area, The threshold for dust concentration. It is the kinetic energy flux per unit time. For reference kinetic flux threshold, This refers to the soil volumetric moisture content or humidity index. This is the baseline value for soil moisture. These are the weight coefficients obtained by training based on historical monitoring data; when At that time, the intelligent early warning processing unit outputs a sandstorm warning signal, in which This is the threshold for triggering the hazard index.

6. A three-dimensional monitoring and early warning system for meteorological gradients of photovoltaic power fields for dust storms according to claim 5, characterized in that, The formula for calculating kinetic flux is: ， in It is the kinetic energy flux per unit time. For counting the kinetic energy pulses of the piezoelectric wind erosion sensor, For single-pulse energy, For measuring time.

7. A three-dimensional monitoring and early warning system for meteorological gradients of photovoltaic power fields for dust storms according to claim 1, characterized in that, The intelligent early warning processing unit fuses and denoises multi-source data using Kalman filtering, multinomial regression, or machine learning algorithms, and dynamically adjusts the hazard index trigger threshold based on the historical training model of the weight coefficients.

8. A three-dimensional monitoring and early warning system for meteorological gradients of photovoltaic power fields for dust storms according to claim 1, characterized in that, The intelligent early warning processing unit identifies the sand-raising wind speed based on the changing trends of particle counts and wind speed data from the piezoelectric wind erosion sensor, and uses the shear wind speed formula. Calculate shear wind speed to dynamically update the reference kinetic flux threshold. and hazard index trigger threshold ,in For shear wind speed, and The wind speeds at two different elevations are as follows: and To correspond to the installation height of the piezoelectric wind erosion sensor, is the von Kármán constant.

9. A three-dimensional monitoring and early warning system for meteorological gradients of photovoltaic power fields and dust storms according to claim 1, characterized in that, The aerial mobile detection unit cruises at at least three different altitudes above the photovoltaic panel along a preset route. Its sensors include particle size measurement or laser scattering dust meters, which are used to generate a three-dimensional distribution of dust concentration and fuse it with ground concentration data.

10. A method for three-dimensional monitoring and early warning of meteorological gradients in photovoltaic power fields for dust storms, characterized in that, The photovoltaic power plant dust meteorological gradient three-dimensional monitoring and early warning system as described in any one of claims 1-9 specifically includes the following steps: Gradient micrometeorological monitoring devices are deployed in different areas of the photovoltaic power plant to collect data on wind speed, wind direction, temperature and humidity, air pressure, precipitation, radiation, soil temperature and humidity, and electrical conductivity. At least two piezoelectric wind erosion sensors at different heights are installed near the ground surface of each monitoring device to collect data on the particle count, kinetic energy pulse, and pulse amplitude of sand impacts. Multi-rotor drones were used to cruise at different altitudes above the photovoltaic panels to measure air temperature, humidity, wind speed, wind direction, air pressure, and dust concentration. All sensor data is collected and stored through the data acquisition and processing unit, and the gain of the piezoelectric wind erosion sensor is remotely controlled. In the intelligent early warning processing unit, the collected data is fused and denoised, and the hazard index I is calculated based on weighting coefficients. Dust storm warnings were issued periodically; By analyzing the inflection point of the particle count versus wind speed curve of a piezoelectric wind erosion sensor, the sand-raising wind speed is identified, and the reference kinetic energy flux threshold is updated by combining it with shear wind speed calculation. and hazard threshold And the sensor sensitivity is adjusted via remote gain control.

Images