Environment evaluation system and evaluation method based on real-time air monitoring data

Abstract

The invention belongs to the technical field of environment monitoring and data analysis, and discloses an environment evaluation system and method based on real-time air monitoring data, and the system comprises an air quality monitoring module which employs a nanometer gas sensitive material sensor array to detect measured pollutants in air, and collects real-time wind field data; combining an unmanned aerial vehicle swarm path planning algorithm, dynamically adjusting the flight path of the unmanned aerial vehicle under the guidance of real-time wind field data, and collecting real-time air monitoring data; the pollution concentration analysis module is used for compensating air monitoring data in a blind area of an acquisition path of the unmanned aerial vehicle through a multi-band sound wave diffraction compensation algorithm so as to generate a continuously distributed pollutant concentration probability field; carrying out space gradient calculation on the pollutant concentration probability field to obtain a pollutant concentration gradient; and the scientificity and effectiveness of environment management are improved, the continuous improvement of the environment quality is promoted, manual intervention is reduced, and the working efficiency and accuracy are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of environmental monitoring and data analysis, and more particularly to an environmental assessment system and assessment method based on real-time air monitoring data. Background Art

[0002] Patent publication number CN117851900A discloses an atmospheric environment control system based on the Internet of Things (IoT), comprising an intelligent sensor network, a data preprocessing module, an environmental simulation and prediction module, a central control unit, an air purification execution unit, a resource management module, and an environmental impact assessment module. The intelligent sensor network is used to monitor pollutants in the atmosphere in real time; the data preprocessing module receives data from the intelligent sensor network and performs preliminary processing on it; the environmental simulation and prediction module simulates the current environmental state and predicts future pollution trends; and the central control unit determines the optimal air purification strategy and directs the coordinated operation of other modules. Through real-time, precise environmental monitoring, intelligent resource management, and comprehensive environmental impact assessment, the system significantly improves the efficiency and accuracy of air quality control, while optimizing energy utilization and reducing negative environmental impacts.

[0003] The existing environmental assessment systems and methods have the following main problems: Existing fixed-site or single-drone monitoring models struggle to cover large, complex areas and can lead to sampling blind spots in terrain-obstructed or high-altitude environments. Drones are easily affected by wind speed and direction during flight, potentially causing them to deviate from their target, impacting monitoring efficiency and data accuracy. Improper drone path planning can easily lead to flight conflicts, which in turn affect the collection of air monitoring data.

[0004] Existing path planning schemes lack the ability to perceive and respond to dynamic environmental factors such as real-time wind fields, resulting in insufficient sampling efficiency and safety. In complex environments, especially areas with high wind speeds and turbulence, a single path planning algorithm may not be able to adapt to dynamically changing environmental conditions. Traditional path planning algorithms are mostly based on static or relatively simple models and cannot fully consider the interaction of multiple dynamic factors such as wind speed and turbulence intensity. There is a lack of a comprehensive tracing mechanism that integrates particle tracking, concentration gradients, and pollution feature matching, resulting in ambiguous identification of pollution sources. In confined or complex areas, there are risks of path conflicts, flight overlaps, and collisions between multiple drones, and there is a lack of real-time obstacle avoidance strategies. Drones may deviate from the target sampling point during obstacle avoidance, and there is a lack of a unified control mechanism to coordinate obstacle avoidance behavior with mission guidance.

[0005] Traditional methods rely solely on sensor sampling, failing to accurately capture pollutant concentration data in areas shadowed by obstacles or with insufficient path coverage, resulting in spatial distribution gaps in monitoring results. Existing methods often employ fixed or linear absorption models, failing to account for the nonlinear influence of pollutant concentration and frequency on absorption, leading to large inversion errors. Single-frequency or simple feature processing methods struggle to identify the concentration characteristics of multiple pollutants from echoes. Conventional kriging interpolation relies on sparse sampling points, failing to consider the authenticity of data in blind spots, and resulting in limited credibility of the interpolated field.

[0006] In view of this, the present invention proposes an environmental assessment system and assessment method based on real-time air monitoring data to solve the above problems. Summary of the Invention

[0007] In order to overcome the above-mentioned defects of the prior art and to achieve the above-mentioned objectives, the present invention provides the following technical solution: an environmental assessment system based on real-time air monitoring data, comprising: The air quality monitoring module uses a nano-gas-sensitive material sensor array to detect pollutants in the air and collect real-time wind field data. Combined with a drone swarm path planning algorithm, it dynamically adjusts the drone collection path under the guidance of real-time wind field data to collect real-time air monitoring data. The pollution concentration analysis module uses a multi-band acoustic diffraction compensation algorithm to compensate for the air monitoring data in the blind spots of the drone's collection path, thereby generating a continuously distributed pollutant concentration probability field. It also calculates the spatial gradient of the pollutant concentration probability field to obtain the pollutant concentration gradient. The pollution source tracing module uses adaptive grid subdivision technology to automatically encrypt regional grids based on air pollutant concentration gradients, generating pollution distribution heat maps. Based on the pollution distribution heat maps, the air pollution source tracing algorithm dynamically analyzes the flow of pollutants. It also uses a preset pollution emission characteristic database to trace the pollution sources in different regions and obtain pollution source tracing data. The regional environmental assessment module collects population health data, evaluates the environment of different regions based on pollution source tracing data and population health data, and predicts environmental quality scores. For different regions, based on the predicted environmental quality scores, it determines whether the environmental quality of the region meets the standards. The evaluation response module monitors the environmental quality of the area every n periods of time and generates an environmental compliance report if the environmental quality of the area meets the standards. If the environmental quality of the area does not meet the standards, an early warning report is automatically generated and sent to the relevant management department.

[0008] Preferably, the method for detecting pollutants in the air and collecting real-time wind field data includes: Set up a drone platform, deploy a nano-gas-sensitive material sensor array on the drone platform, which is integrated on one or m micro-control flow chips, detect trace pollutants in the air, and perform analog-to-digital conversion to generate trace pollutant data with timestamp and location information. Collect real-time wind field data through a hot film anemometer. The real-time wind field data includes wind speed, wind direction, turbulence intensity and vertical wind shear.

[0009] Preferably, the method for collecting real-time air monitoring data includes: Preset a target monitoring area for collecting air monitoring data, deploy N drones in the target monitoring area. Preset that when there is no wind, the speed of the drone flying towards the target point in the target monitoring area is V1, and the wind speed at the position of the target point in the target monitoring area is V2. Then, when there is wind, the actual flight speed of the drone is V3 = V1 + V2; when there is wind, preset the ideal flight speed of the drone as V. Subtract the wind speed V2 at the position of the target point in the target monitoring area from the preset ideal flight speed V of the drone to obtain the speed V4 that the drone needs to correct. Send an adjustment instruction through the drone platform terminal, continuously adjust the speed V4 that the drone needs to correct. Subtract the actual flight speed V3 of the drone in windy conditions from the speed V4 that the drone needs to correct to obtain a correction deviation. Preset a correction deviation threshold. When the correction deviation is less than or equal to the preset correction deviation threshold, send a stop instruction through the drone platform terminal to stop the correction of the drone speed. Take the wind speed, turbulence intensity and vertical wind shear at the position of the target point in the target monitoring area as path cost function factors, and the drone calculates the path cost function of the flight path in real time. Combine the behaviors of worker bees, observer bees and scout bees in the drone swarm path planning algorithm to update or replan the path, and form an optimal coverage path for the target monitoring area. Preset the position coordinates of drone i as (xi, yi, zi), and the position coordinates of drone j as (xj, yj, zj). Calculate the Euclidean distance dij between drone i and drone j. Preset the safety distance between drones as ds. If dij < ds, it is determined that there is a collision risk between drone i and drone j, and the flight path of the drone needs to be adjusted. Dynamically adjust the safety distance between preset drones through the safety distance adjustment formula. The safety distance adjustment formula is ; where represents the safety distance between drones after dynamic adjustment; represents the battery power attenuation influence coefficient; represents the maximum battery power of the drone; represents the current battery power of the drone; Each UAV is regarded as a particle with repulsive characteristics, and the target points in the target monitoring area are regarded as source points with gravitational characteristics; when dij < dr, the cooperative control mechanism of repulsive force and gravitational force is triggered, and repulsive adjustment instructions are applied according to the relative direction between UAVs to guide the corresponding UAVs away from the collision path; and a guiding force in the direction of the target point is applied to the UAV based on the position of the target point in the target monitoring area to guide the UAV to fly towards the target monitoring area; the repulsive force and the guiding force are superimposed to form a resultant force direction, and the flight path of the corresponding UAV is dynamically adjusted according to the resultant force direction; real-time air monitoring data is collected, and the real-time air monitoring data includes pollutant measurement data, air temperature and humidity data, and wind speed and direction data in the target monitoring area.

[0010] Preferably, the method for generating the continuously distributed pollutant concentration probability field includes: For the blind area of the UAV acquisition path, a multi-band acoustic wave transmitting and receiving module is installed on the UAV platform, and multi-frequency acoustic wave signals are transmitted towards the blind area of the UAV acquisition path during the flight of the UAV; by receiving the echo signals after the acoustic waves diffract at the edge of the obstacle and in the air medium; extracting the characteristics of the echo signals, and the characteristics of the echo signals include the time delay of the echo signal, the intensity of the echo signal, and the phase change. For different pollutant concentrations and pollutant types, calculate the pollutant diffraction response function of the multi-frequency acoustic wave in each frequency band, combine the pollutant diffraction response function and the characteristics of the echo signal, and use the least squares method for inversion to estimate the pollutant concentration in the blind area of the UAV acquisition path. The Kriging interpolation method is used to extend the inversion result to the entire target monitoring area to generate a continuously distributed pollutant concentration probability field.

[0011] Preferably, the method for obtaining the pollutant concentration gradient includes; The target monitoring area is divided into multiple grid cells, and each grid cell corresponds to an estimated pollutant concentration; the pollutant concentration of any position point in each grid cell is compared with the pollutant concentration of any position point in its surrounding adjacent grid cells, and the change amount of the pollutant concentration in the horizontal direction, vertical direction, and height direction at this position point is extracted. The change amounts of the concentration in the horizontal direction, vertical direction, and height direction are combined into a vector to represent the spatial gradient of the pollutant concentration at this position point; by extracting the change amounts of the pollutant concentration at all grid cell position points, the pollutant concentration gradient of the pollutant concentration probability field is obtained.

[0012] Preferably, the method for generating the pollution distribution heat map includes: A pollutant concentration gradient threshold is preset. If the pollutant concentration gradient of any grid cell in the target monitoring area is greater than the preset pollutant concentration gradient threshold, it is determined that the grid cell needs to be encrypted and subdivided; if the pollutant concentration gradient of any grid cell in the target monitoring area is less than or equal to the preset pollutant concentration gradient threshold, it is determined that the grid cell does not need to be encrypted and subdivided; The grid cells that need to be refined and subdivided are divided into different sub-grid cells; the pollutant concentration is estimated again for each sub-grid cell; and the estimated pollutant concentrations of all grid cells are projected onto a map, using color depth to express different pollutant concentrations to obtain a pollution distribution heat map.

[0013] Preferably, the method for obtaining the pollution source tracing data includes: The preset pollution emission characteristic database is composed of the emission characteristic vectors of the pollution sources, which include pollutant type data, pollutant emission intensity data and pollutant emission time data. The Lagrangian particle backtracking method is used to regard the pollutants in the pollution distribution heat map as particles, and the positions of their suspected source points are traced back. The particles are backtracked by the fourth-order Runge-Kutta method to obtain the entire particle trajectory. The pollutant type, pollutant concentration and pollution time at the starting point of the particle trajectory are extracted to form a pollutant characteristic vector. The pollutant characteristic vector is matched with the emission characteristic vector of the pollution source in the preset pollution emission characteristic database by cosine similarity to obtain the pollution source matching result, and the pollution source matching result is sorted in descending order according to the similarity to obtain the pollution tracing data.

[0014] Preferably, the method for obtaining the environmental quality score includes: Construct and train an environmental quality assessment model, which includes an input layer, a GRU layer, a fully connected layer, and an output layer. The input layer of the environmental quality assessment model is used to input historical pollution source tracing data and population health data, and the output layer of the model is used to output the environmental quality score, using the identity function as the activation function. The environmental quality assessment model is a gated recurrent unit model, and the root mean square error is used as the loss function of the model. The current pollution source tracing data and population health data are input into the trained environmental quality assessment model to obtain the environmental quality score.

[0015] Preferably, the method for determining whether the environmental quality of the area meets the standards includes: Preset the environmental quality score threshold, select any area, and compare the predicted environmental quality score with the preset environmental quality score threshold. If the predicted environmental quality score is greater than or equal to the preset environmental quality score threshold, the environmental quality of the area is judged to be up to standard; if the predicted environmental quality score is less than the preset environmental quality score threshold, the environmental quality of the area is judged to be not up to standard.

[0016] Environmental assessment methods based on real-time air monitoring data include: S1. Use a nano-gas-sensitive material sensor array to detect pollutants in the air and collect real-time wind field data. Combined with a drone swarm path planning algorithm, the drone collection path is dynamically adjusted under the guidance of real-time wind field data to collect real-time air monitoring data. S2. Compensate the air monitoring data within the blind spots of the drone's collection path using a multi-band acoustic diffraction compensation algorithm, thereby generating a continuously distributed pollutant concentration probability field. Calculate the spatial gradient of the pollutant concentration probability field to obtain the pollutant concentration gradient. S3. Using adaptive grid subdivision technology, we automatically refine regional grids based on air pollutant concentration gradients to generate pollution distribution heat maps. Based on these heat maps, we dynamically analyze the flow of pollutants using an air pollution source tracing algorithm. We also use a pre-set pollution emission characteristic database to trace pollution sources in different regions and generate pollution source tracing data. S4. Collect population health data, evaluate the environment of different regions based on pollution source tracing data and population health data, and predict environmental quality scores. For different regions, determine whether the environmental quality of the region meets the standards based on the predicted environmental quality scores; S5. If the environmental quality of the area meets the standard, the environmental quality of the area will be monitored every n periods of time and an environmental compliance report will be generated; if the environmental quality of the area does not meet the standard, an early warning report will be automatically generated and sent to the relevant management department.

[0017] Compared with the prior art, the present invention has the following beneficial effects: By calculating and adjusting drone flight speed, direction, and path planning in real time, the system ensures efficient flight and accurate arrival at target locations despite changing factors such as wind speed. This significantly improves the accuracy of air quality monitoring and the coverage of monitoring points. By introducing a collision avoidance algorithm and a safe distance adjustment formula, collisions between multiple drones during flight are avoided, ensuring the safety of multi-drone coordinated flight. This approach not only enhances flight safety but also reduces the chance of flight path overlap, thereby increasing efficiency.

[0018] The flight path is optimized in real time based on environmental factors such as wind speed and turbulence, and the path is updated in combination with the swarm algorithm. It can flexibly adapt to environmental changes and maintain efficient execution of monitoring tasks. This flexible path planning and adjustment method enhances the adaptability and work efficiency of drones in complex environments. In windy conditions, the flight speed of the drone is dynamically corrected to overcome the flight deviation caused by wind speed changes, thereby ensuring that the drone can always fly according to the predetermined path, effectively improving the quality of monitoring data. The introduction of a collaborative control mechanism of repulsive force and guiding force enables the drone to dynamically adjust according to the surrounding environment, the direction of the target point, and the relative position of other drones. This not only avoids collisions, but also guides the drone to maintain efficient path planning and task execution; using adaptive grid encryption and pollution concentration gradient extraction methods, combined with heat map generation and pollution feature database matching, high-resolution pollution migration analysis and precise tracing are achieved; Through a multi-frequency acoustic diffraction sensing mechanism, indirect estimation of previously unmonitorable areas, such as blind spots and areas behind obstacles, is achieved, significantly improving the spatial continuity of monitoring data. Absorption coefficient constraints limit extreme or non-physical results, improving the credibility and stability of pollution concentration estimates. The system supports a multi-band, multi-parameter acoustic modeling framework that adapts to different pollutant types and their acoustic response characteristics, demonstrating excellent versatility and scalability. The continuously distributed probability field provides a more granular description of pollutant distribution, serving as high-quality input for subsequent models such as pollution diffusion simulation and exposure risk analysis. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] Figure 1 Schematic diagram of the structure of the environmental assessment system based on real-time air monitoring data of the present invention; Figure 2 Schematic diagram of the process of the environmental assessment method based on real-time air monitoring data of the present invention; Figure 3 This is a flow chart of the method for generating a pollution distribution heat map provided by the present invention. DETAILED DESCRIPTION

[0020] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0021] Example 1 See also Figure 1 and Figure 3 As shown, this embodiment 1 further illustrates the environmental assessment system based on real-time air monitoring data proposed by the present invention, including: In recent years, with the rapid development of industrialization and urbanization, air pollution has become increasingly serious, becoming a major environmental issue affecting human health and ecological security. Traditional air quality monitoring relies primarily on fixed ground-based stations. These stations are limited in number, spatial distribution, and real-time performance, making them unable to meet the demand for large-scale, detailed, and dynamic air quality information.

[0022] To address this issue, a growing number of researchers are introducing drone platforms as a new mobile monitoring tool. Drones offer advantages such as flexible deployment, high maneuverability, and low cost. They can collect three-dimensional air pollution data in complex terrain, hazardous areas, or areas traditionally blinded by monitoring, providing higher spatial resolution and more timely monitoring data. Drones have become a vital addition to the field of environmental monitoring.

[0023] On the other hand, with the development of sensor technology and data communication technology, micro-sensors that can obtain real-time environmental parameters such as air pollutant concentration, temperature and humidity, wind speed and direction have become increasingly popular, making it possible to achieve dynamic assessment and tracing of air pollution.

[0024] However, in the prior art, the following major problems still exist: Difficulties in coordinating and scheduling multiple drones: Traditional path planning struggles to handle multi-objective scheduling challenges such as complex wind field interference, collision avoidance requirements, and power constraints, often resulting in low monitoring efficiency or safety risks. Inaccurate pollutant source tracking: The lack of accurate pollutant migration modeling and particle tracing mechanisms makes it difficult to identify pollution sources with high confidence; Lack of real-time response capabilities: Most environmental assessment methods rely on batch data analysis and cannot achieve real-time identification and response to pollution incidents; Separation of path planning and collision avoidance control: Existing systems often separate path planning and obstacle avoidance control, making it impossible to achieve overall optimal path decision-making in dynamic environments. The pollution source tracing and characteristic database are loosely matched: the lack of a systematic design of the pollution source emission characteristic database leads to low pollution identification accuracy and a lack of basis for tracing results.

[0025] Therefore, in order to effectively solve the above problems, the present invention proposes an environmental assessment system based on real-time air monitoring data, comprising: The air quality monitoring module uses a nano-gas-sensitive material sensor array to detect pollutants in the air and collect real-time wind field data. Combined with a drone swarm path planning algorithm, it dynamically adjusts the drone collection path under the guidance of real-time wind field data to collect real-time air monitoring data. The pollution concentration analysis module uses a multi-band acoustic diffraction compensation algorithm to compensate for the air monitoring data in the blind spots of the drone's collection path, thereby generating a continuously distributed pollutant concentration probability field. It also calculates the spatial gradient of the pollutant concentration probability field to obtain the pollutant concentration gradient. The pollution source tracing module uses adaptive grid subdivision technology to automatically encrypt regional grids based on air pollutant concentration gradients, generating pollution distribution heat maps. Based on the pollution distribution heat maps, the air pollution source tracing algorithm dynamically analyzes the flow of pollutants. It also uses a preset pollution emission characteristic database to trace the pollution sources in different regions and obtain pollution source tracing data. The regional environmental assessment module collects population health data, evaluates the environment of different regions based on pollution source tracing data and population health data, and predicts environmental quality scores. For different regions, based on the predicted environmental quality scores, it determines whether the environmental quality of the region meets the standards. Among them, population health data refers to indicators that can represent the health level of the population in the region, such as the overall health rate (obtained as the percentage of healthy people to the total population) or the proportion of healthy people in diseases affected by environmental factors. The collection method can be retrieved from the database of the CDC in the region, from the hospital data center, or through questionnaire surveys.

[0026] The evaluation response module monitors the environmental quality of the area every n periods of time and generates an environmental compliance report if the environmental quality of the area meets the standards. If the environmental quality of the area does not meet the standards, an early warning report is automatically generated and sent to the relevant management department.

[0027] Methods for measuring pollutants and collecting real-time wind data include: A drone platform is deployed, on which a nano-gas-sensitive material sensor array is deployed. These sensors are integrated into one or more micro-control flow chips to detect airborne pollutants, including nitrogen oxides, sulfur dioxide, carbon monoxide, particulate matter, ozone, volatile organic compounds, formaldehyde, hydrogen sulfide, chlorine, and heavy metals. Analog-to-digital conversion is performed to generate pollutant data with timestamps and location information. Real-time wind field data, including wind speed, wind direction, turbulence intensity, and vertical wind shear, is collected and detected using a hot-film anemometer. Vertical wind shear refers to the rate of change of wind speed or direction between different altitudes—that is, the change in wind speed and / or direction with altitude.

[0028] Methods for collecting real-time air monitoring data include: Preset a target monitoring area for collecting air monitoring data, deploy N drones in the target monitoring area. Preset that when there is no wind, the flying speed of the drone towards the target point in the target monitoring area is V1, and the wind speed at the position of the target point in the target monitoring area is V2. Then, in the case of wind, the actual flying speed of the drone is V3 = V1 + V2. It should be noted that the wind speed at the position of the target point in the target monitoring area is directional here. The direction of V1 and V2 may be the same or opposite. If they are opposite, then V3 = V1 + (-V2). To ensure that the drone can still fly towards the target point in the case of wind, in the case of wind, preset the ideal flying speed of the drone as V. Subtract the wind speed V2 at the position of the target point in the target monitoring area from the preset ideal flying speed V of the drone to obtain the speed V4 that the drone needs to correct. Send an adjustment instruction through the drone platform terminal, continuously adjust the speed V4 that the drone needs to correct. Subtract the actual flying speed V3 of the drone in the case of wind from the speed V4 that the drone needs to correct to obtain a correction deviation. Preset a correction deviation threshold. When the correction deviation is less than or equal to the preset correction deviation threshold, send a stop instruction through the drone platform terminal to stop the correction of the drone speed. Take the wind speed, turbulence intensity, and vertical wind shear at the position of the target point in the target monitoring area as path cost function factors, and calculate the path cost function of the drone's flight path in real time. Combine the behaviors of worker bees, observer bees, and scout bees in the drone swarm path planning algorithm to update or replan the path, and form an optimal coverage path for the target monitoring area. When multiple drones fly in coordination, problems such as path overlap and collision may occur, especially in a narrow monitoring area or a complex environment. To solve the possible collision problems between multiple drones, introduce a collision avoidance algorithm to adjust the flight path of the drones in real time. Preset the position coordinates of drone i as (xi, yi, zi), and the position coordinates of drone j as (xj, yj, zj). Calculate the Euclidean distance dij between drone i and drone j. Preset the safe distance between drones as ds. If dij < ds, it is determined that there is a collision risk between drone i and drone j, and the flight path of the drone needs to be adjusted. Dynamically adjust the safe distance between preset drones through the safe distance adjustment formula. The safe distance adjustment formula is ; where represents the safe distance between drones after dynamic adjustment; represents the battery power attenuation influence coefficient; represents the maximum battery power of the drone; represents the current battery power of the drone; It should be noted that at low battery levels, the thrust output and response time of the drone decrease: the motor may operate at reduced power and may switch to an energy-saving or return-to-home mode; high-speed emergency obstacle avoidance operations are restricted and evasive maneuvers cannot be made in a timely manner. Therefore, a larger safety buffer (i.e., increasing the minimum safe distance between drones) is required to prevent collisions. The safe distance should not be a fixed value and should be related to the flight state; battery level, as one of the most critical state variables, is a reasonable reference factor for dynamically adjusting safety redundancy.

[0029] Each drone is regarded as a particle with repulsive characteristics, and the target points in the target monitoring area are regarded as source points with gravitational characteristics; when dij < dr, the collaborative control mechanism of repulsive force and gravitational force is triggered, and a repulsive adjustment instruction is applied according to the relative direction between the drones to guide the corresponding drone away from the collision path; and a guiding force in the direction of the target point is applied to the drone based on the position of the target point in the target monitoring area to guide the drone to fly towards the target monitoring area; the repulsive force and the guiding force are superimposed to form a resultant force direction, and the flight path of the corresponding drone is dynamically adjusted according to the resultant force direction; real-time air monitoring data is collected, and the real-time air monitoring data includes pollutant measurement data, air temperature and humidity data, and wind speed and direction data in the target monitoring area.

[0030] The following problems existing in the prior art are solved: In the actual monitoring environment, the wind speed has directionality and time-variability, and it is very easy for the drone to deviate from the target path when flying in the wind field; this may cause the drone to deviate from the target point during actual flight, thus affecting the monitoring efficiency and data accuracy. When multiple drones fly in coordination, they often face problems such as path overlap and collision, especially in narrow or complex environments. In this case, if the path planning of the drone is improper, it is easy to cause flight conflicts, which in turn affects the collection of monitoring data.

[0031] Traditional path planning methods are mostly static planning, unable to respond to meteorological factors such as wind speed, turbulence, and wind shear in real time, with poor path adaptability, resulting in the failure of the acquisition trajectory or repeated sampling; in complex environments, especially in areas with strong wind speed and turbulence, a single path planning algorithm may not be able to adapt to the dynamically changing environmental conditions. Traditional path planning algorithms are mostly based on static or relatively simple models and cannot comprehensively consider the interaction of multiple dynamic factors such as wind speed and turbulence intensity. In narrow or complex areas, there are risks of path conflicts, flight overlap, and even collision between multiple drones, lacking real-time obstacle avoidance strategies; existing drone obstacle avoidance mostly sets a fixed safety distance and does not consider the impact of battery level changes on flight control capabilities, which may result in obstacle avoidance failure in the low battery state; the drone may deviate from the target sampling point during obstacle avoidance, lacking a unified control mechanism for coordinating obstacle avoidance behavior and task guidance. The path cannot be adaptively updated according to the real-time wind field and drone distribution, resulting in blind spots or redundancy in data collection.

[0032] Advantages compared to existing technologies: By calculating and adjusting drone flight speed, direction, and path planning in real time, the system ensures efficient flight and accurate arrival at target locations despite changing factors such as wind speed. This significantly improves the accuracy of air quality monitoring and the coverage of monitoring points. By introducing a collision avoidance algorithm and a safe distance adjustment formula, collisions between multiple drones during flight are avoided, ensuring the safety of multi-drone coordinated flight. This approach not only enhances flight safety but also reduces the chance of flight path overlap, thereby increasing efficiency.

[0033] Real-time flight path optimization based on environmental factors such as wind speed and turbulence, combined with a swarm algorithm for path updates, allows for flexible adaptation to environmental changes and maintains efficient monitoring mission execution. This flexible path planning and adjustment enhances the drone's adaptability and efficiency in complex environments. In windy conditions, dynamic correction of the drone's flight speed overcomes flight deviations caused by wind speed fluctuations, ensuring the drone consistently follows the planned path and effectively improving the quality of monitoring data. A coordinated control mechanism for repulsive and guiding forces enables the drone to dynamically adjust based on the surrounding environment, the target's direction, and the relative position of other drones. This not only avoids collisions but also guides the drone to maintain efficient path planning and mission execution.

[0034] Methods for generating a continuously distributed pollutant concentration probability field include: To address blind spots in the drone's acquisition path, a multi-band acoustic wave transmitting and receiving module is installed on the drone platform. During flight, multi-frequency acoustic wave signals are transmitted toward the blind spots in the drone's acquisition path. The echo signals are received after the acoustic waves are diffracted by the obstacle edge and the air medium. The echo signal features are extracted, including echo signal delay, echo signal strength, and phase change. For different pollutant concentrations and types, calculate the pollutant diffraction response function of multi-frequency sound waves in each frequency band ;in, Indicates the frequency of the sound wave; Indicates the concentration of any pollutant in the target area; Indicates the length of the sound wave propagation path; represents the absorption coefficient, which describes the attenuation of sound waves due to absorption by pollutants during propagation; It indicates the phase change caused by the change in propagation speed (sound velocity disturbance caused by pollutants) during the propagation of sound waves; Indicates the initial signal amplitude when the sound wave is emitted; represents an imaginary unit; In the process of generating a continuously distributed pollutant concentration probability field based on multi-frequency acoustic inversion, the pollutant's absorption characteristics of acoustic waves are modeled using the absorption coefficient. However, unconstrained absorption coefficients can exhibit unphysical negative values, leading to an "amplification" phenomenon during acoustic wave propagation, violating the fundamental law that acoustic waves gradually attenuate due to energy loss during actual propagation. Furthermore, if the acoustic wave propagation path is long or the absorption coefficient is abnormal, if the absorption coefficient is too large, the exponential term will quickly approach zero, easily leading to floating-point underflow. Conversely, if the absorption coefficient is too small or even negative, the signal strength will be abnormally amplified, causing numerical instability.

[0035] Furthermore, unconstrained absorption coefficients directly impact the accuracy of pollutant concentration inversion. Because the mapping between echo characteristics and pollutant concentrations relies on the accuracy of the absorption model, distorted absorption coefficients can cause concentration estimates to deviate from actual values, significantly increasing errors. Further kriging interpolation calculations based on this assumption will expand the pollutant concentration probability field across the entire target area based on the erroneous initial inversion results, rendering the monitoring map unreliable.

[0036] The absorption coefficient is constrained by the absorption coefficient constraint formula, which is: ;in, represents the maximum absorption coefficient; Indicates the frequency growth influence coefficient, which adjusts the response speed of the absorption coefficient to the frequency; Indicates the concentration growth influence coefficient, which adjusts the response speed of the absorption coefficient to the pollutant concentration; It should be noted that the design logic of the absorption coefficient constraint formula is based on the actual mechanism of sound wave propagation in a polluted environment and the mathematical expression habits of typical natural phenomena; the absorption coefficient increases with increasing frequency, and when sound waves propagate in the air, high-frequency sound waves are more easily absorbed by the medium (including polluted gases). Therefore, the absorption coefficient increases with increasing frequency. The higher the concentration of pollutants, the more frequent the collisions with gas molecules during sound wave propagation, and the greater the energy dissipation, so the degree of absorption will also increase with increasing concentration. The absorption coefficient constraint formula is a common growth-type saturation function that can describe rapid growth in the initial stage; as the input variables (frequency, concentration) continue to increase, the growth tends to be slow, and eventually approaches an upper limit value; The present invention solves the following technical problems existing in the prior art: Traditional methods rely solely on sensor sampling and are unable to accurately obtain pollutant concentration data in areas shadowed by obstacles or with insufficient path coverage, resulting in spatial distribution gaps in monitoring results. Existing methods often use fixed or linear absorption models, which cannot reflect the nonlinear influence of pollutant concentration and frequency on the absorption effect, resulting in large inversion errors. Single-frequency or simple feature processing methods make it difficult to identify the concentration characteristics of multiple pollutants from echoes. Conventional kriging interpolation is based on sparse sampling points and does not consider the authenticity of blind area data, resulting in limited credibility of the interpolated field. Advantages compared to existing technologies: Through a multi-frequency acoustic diffraction sensing mechanism, indirect estimation of previously unmonitorable areas, such as blind spots and areas behind obstacles, is achieved, significantly improving the spatial continuity of monitoring data. Absorption coefficient constraints limit extreme or non-physical results, improving the credibility and stability of pollution concentration estimates. The system supports a multi-band, multi-parameter acoustic modeling framework that adapts to different pollutant types and their acoustic response characteristics, demonstrating excellent versatility and scalability. The continuously distributed probability field provides a more granular description of pollutant distribution, serving as high-quality input for subsequent models such as pollution diffusion simulation and exposure risk analysis.

[0037] Combining the pollutant diffraction response function and echo signal characteristics, the least squares method is used to invert and estimate the pollutant concentration in the blind spot of the UAV collection path. The Kriging interpolation method is used to extend the inversion results to the entire target monitoring area to generate a continuously distributed pollutant concentration probability field.

[0038] Methods for obtaining pollutant concentration gradients include: The target monitoring area is divided into multiple grid cells, each of which corresponds to an estimated pollutant concentration; the pollutant concentration of any point in each grid cell is compared with that of any point in its surrounding adjacent grid cells, and the pollutant concentration changes at the point in the horizontal, vertical, and height directions are extracted. The concentration changes in the horizontal, vertical, and height directions are combined into a vector to represent the spatial gradient of the pollutant concentration at the point; by extracting the pollutant concentration changes at all grid cell points, the pollutant concentration gradient of the pollutant concentration probability field is obtained. For example, ;in, Indicates along The change of pollutant concentration in the horizontal direction; Indicates the horizontal coordinate of any position in the target monitoring area; Indicates the horizontal distance between two adjacent grid cells; and so on. and .

[0039] The methods for generating pollution distribution heat maps include: A pollutant concentration gradient threshold is preset. If the pollutant concentration gradient of any grid cell in the target monitoring area is greater than the preset pollutant concentration gradient threshold, it is determined that the grid cell needs to be encrypted and subdivided; if the pollutant concentration gradient of any grid cell in the target monitoring area is less than or equal to the preset pollutant concentration gradient threshold, it is determined that the grid cell does not need to be encrypted and subdivided; The grid cells that need to be refined and subdivided are divided into different sub-grid cells; the pollutant concentration is estimated again for each sub-grid cell; and the estimated pollutant concentrations of all grid cells are projected onto a map, using color depth to express different pollutant concentrations to obtain a pollution distribution heat map.

[0040] Methods for obtaining pollution source tracing data include: The preset pollution emission characteristic database is composed of the emission characteristic vectors of the pollution sources, which include pollutant type data, pollutant emission intensity data and pollutant emission time data. The Lagrangian particle backtracking method is used to regard the pollutants in the pollution distribution heat map as particles, and the positions of their suspected source points are traced back. The particles are backtracked by the fourth-order Runge-Kutta method to obtain the entire particle trajectory. The pollutant type, pollutant concentration and pollution time at the starting point of the particle trajectory are extracted to form a pollutant characteristic vector. The pollutant characteristic vector is matched with the emission characteristic vector of the pollution source in the preset pollution emission characteristic database by cosine similarity to obtain the pollution source matching result, and the pollution source matching result is sorted in descending order according to the similarity to obtain the pollution tracing data.

[0041] Methods for obtaining environmental quality scores include: Construct and train an environmental quality assessment model, which includes an input layer, a GRU layer, a fully connected layer, and an output layer. The input layer of the environmental quality assessment model is used to input historical pollution source tracing data and population health data, and the output layer of the model is used to output the environmental quality score, using the identity function as the activation function. The environmental quality assessment model is a gated recurrent unit model, and the root mean square error is used as the loss function of the model. The current pollution source tracing data and population health data are input into the trained environmental quality assessment model to obtain the environmental quality score.

[0042] Methods for determining whether the environmental quality of the area meets the standards include: Preset the environmental quality score threshold, select any area, and compare the predicted environmental quality score with the preset environmental quality score threshold. If the predicted environmental quality score is greater than or equal to the preset environmental quality score threshold, the environmental quality of the area is judged to be up to standard; if the predicted environmental quality score is less than the preset environmental quality score threshold, the environmental quality of the area is judged to be not up to standard.

[0043] The preset environmental quality score threshold is set by the staff by taking the average of multiple environmental quality scores as the preset environmental quality score threshold; similarly, the preset pollutant concentration gradient threshold and the preset correction deviation threshold are set.

[0044] This embodiment, through real-time calculation and adjustment of drone flight speed, direction, and path planning, ensures that drones can fly efficiently and accurately reach their target locations despite the influence of factors such as wind speed. This significantly improves the accuracy of air quality monitoring and the coverage of monitoring points. By introducing a collision avoidance algorithm and a safe distance adjustment formula, collisions between multiple drones during flight are avoided, ensuring the safety of multi-drone coordinated flight. This approach not only improves flight safety but also reduces the chance of flight path overlap, thereby improving efficiency.

[0045] The flight path is optimized in real time based on environmental factors such as wind speed and turbulence, and the path is updated in combination with the swarm algorithm. It can flexibly adapt to environmental changes and maintain efficient execution of monitoring tasks. This flexible path planning and adjustment method enhances the adaptability and work efficiency of drones in complex environments. In windy conditions, the flight speed of the drone is dynamically corrected to overcome the flight deviation caused by wind speed changes, thereby ensuring that the drone can always fly according to the predetermined path, effectively improving the quality of monitoring data. The introduction of a collaborative control mechanism of repulsive force and guiding force enables the drone to dynamically adjust according to the surrounding environment, the direction of the target point, and the relative position of other drones. This not only avoids collisions, but also guides the drone to maintain efficient path planning and task execution; using adaptive grid encryption and pollution concentration gradient extraction methods, combined with heat map generation and pollution feature database matching, high-resolution pollution migration analysis and precise tracing are achieved; Through a multi-frequency acoustic diffraction sensing mechanism, indirect estimation of previously unmonitorable areas, such as blind spots and areas behind obstacles, is achieved, significantly improving the spatial continuity of monitoring data. Absorption coefficient constraints limit extreme or non-physical results, improving the credibility and stability of pollution concentration estimates. The system supports a multi-band, multi-parameter acoustic modeling framework that adapts to different pollutant types and their acoustic response characteristics, demonstrating excellent versatility and scalability. The continuously distributed probability field provides a more granular description of pollutant distribution, serving as high-quality input for subsequent models such as pollution diffusion simulation and exposure risk analysis.

[0046] Example 2 See also Figure 2 As shown, for the parts not described in detail in this embodiment, please refer to the description of Example 1. An environmental assessment method based on real-time air monitoring data is provided, including: S1. Use a nano-gas-sensitive material sensor array to detect pollutants in the air and collect real-time wind field data. Combined with a drone swarm path planning algorithm, the drone collection path is dynamically adjusted under the guidance of real-time wind field data to collect real-time air monitoring data. S2. Compensate the air monitoring data within the blind spots of the drone's collection path using a multi-band acoustic diffraction compensation algorithm, thereby generating a continuously distributed pollutant concentration probability field. Calculate the spatial gradient of the pollutant concentration probability field to obtain the pollutant concentration gradient. S3. Using adaptive grid subdivision technology, we automatically refine regional grids based on air pollutant concentration gradients to generate pollution distribution heat maps. Based on these heat maps, we dynamically analyze the flow of pollutants using an air pollution source tracing algorithm. We also use a pre-set pollution emission characteristic database to trace pollution sources in different regions and generate pollution source tracing data. S4. Collect population health data, evaluate the environment of different regions based on pollution source tracing data and population health data, and predict environmental quality scores. For different regions, determine whether the environmental quality of the region meets the standards based on the predicted environmental quality scores; S5. If the environmental quality of the area meets the standard, the environmental quality of the area will be monitored every n periods of time and an environmental compliance report will be generated; if the environmental quality of the area does not meet the standard, an early warning report will be automatically generated and sent to the relevant management department.

[0047] Since the electronic device introduced in this embodiment is an electronic device used to implement the environmental assessment system and assessment method based on real-time air monitoring data in the embodiment of the present application, based on the environmental assessment system and assessment method based on real-time air monitoring data introduced in the embodiment of the present application, those skilled in the art can understand the specific implementation of the electronic device of this embodiment and its various variations, so how the electronic device implements the method in the embodiment of the present application will not be described in detail here. As long as those skilled in the art implement the electronic device used in the environmental assessment system and assessment method based on real-time air monitoring data in the embodiment of the present application, they all fall within the scope of protection of this application.

[0048] The above formulas are all dimensionless and numerical calculations. The formulas are obtained by collecting a large amount of data and performing software simulation to obtain the most recent real situation. The preset parameters and thresholds in the formulas are set by technicians in this field according to actual conditions.

[0049] The above description is merely a preferred embodiment of the present invention. The scope of protection of the present invention is not limited to the above embodiment. All technical solutions based on the principles of the present invention are within the scope of protection of the present invention. It should be noted that for users of ordinary skill in the art, various improvements and modifications that do not depart from the principles of the present invention should also be considered within the scope of protection of the present invention.

Claims

1. An environmental assessment system based on real-time air monitoring data, characterized by: include: Air quality monitoring module, which uses a nano-gas-sensitive material sensor array to detect pollutants in the air and collect real-time wind field data; Combined with the drone swarm path planning algorithm, the drone collection path is dynamically adjusted under the guidance of real-time wind field data to collect real-time air monitoring data; The pollution concentration analysis module uses a multi-band acoustic diffraction compensation algorithm to compensate for the air monitoring data in the blind spots of the drone's collection path, thereby generating a continuously distributed pollutant concentration probability field. It also calculates the spatial gradient of the pollutant concentration probability field to obtain the pollutant concentration gradient. The pollution source tracing module uses adaptive grid subdivision technology to automatically encrypt regional grids based on air pollutant concentration gradients, generating pollution distribution heat maps. Based on the pollution distribution heat maps, the air pollution source tracing algorithm dynamically analyzes the flow of pollutants. It also uses a preset pollution emission characteristic database to trace the pollution sources in different regions and obtain pollution source tracing data. The regional environmental assessment module collects population health data, evaluates the environment of different regions based on pollution source tracing data and population health data, and predicts environmental quality scores. For different regions, based on the predicted environmental quality scores, it determines whether the environmental quality of the region meets the standards. The evaluation response module monitors the environmental quality of the area every n periods of time and generates an environmental compliance report if the environmental quality of the area meets the standards. If the environmental quality of the area does not meet the standards, an early warning report is automatically generated and sent to the relevant management department.

2. The environmental assessment system based on real-time air monitoring data according to claim 1, characterized in that: The method for detecting pollutants in the air and collecting real-time wind field data includes: A drone platform is set up, and a nano-gas-sensitive material sensor array is deployed on the drone platform. The sensor array is integrated on one or m micro-control flow chips to detect pollutants in the air and perform analog-to-digital conversion to generate pollutant data with timestamp and location information. Real-time wind field data is obtained through hot-film anemometer collection and detection. The real-time wind field data includes wind speed, wind direction, turbulence intensity and vertical wind shear.

3. The environmental assessment system based on real-time air monitoring data according to claim 2, characterized in that: The method for collecting real-time air monitoring data includes: A target monitoring area for collecting air monitoring data is preset, and N drones are deployed in the target monitoring area. When there is no wind, the speed of the drone flying towards the target point in the target monitoring area is preset to be V1, and the wind speed at the target point in the target monitoring area is preset to be V2. Then, when there is wind, the actual flight speed of the drone is V3=V1+V2. When there is wind, the ideal flight speed of the drone is preset to be V. The wind speed V2 at the target point in the target monitoring area is subtracted from the ideal flight speed V to obtain the corrected speed V4 of the drone. An adjustment command is issued through the UAV platform terminal to continuously adjust the speed V4 of the UAV that needs to be corrected. The speed V4 that needs to be corrected of the UAV is subtracted from the actual flight speed V3 of the UAV in windy conditions to obtain a correction deviation. A correction deviation threshold is preset. When the correction deviation is less than or equal to the preset correction deviation threshold, a stop command is issued through the UAV platform terminal to stop correcting the UAV speed. Taking the wind speed, turbulence intensity, and vertical wind shear at the target point position in the target monitoring area as path cost function factors, the UAV calculates the path cost function of the flight path in real time, and combines the behaviors of worker bees, observer bees, and scout bees in the UAV swarm path planning algorithm to update or replan the path, forming an optimal coverage path for the target monitoring area; Preset the position coordinates of UAV i as (xi, yi, zi), and the position coordinates of UAV j as (xj, yj, zj). Calculate the Euclidean distance dij between UAV i and UAV j. Preset the safety distance between UAVs as ds. If dij < ds, it is determined that there is a collision risk between UAV i and UAV j, and the flight path of the UAV needs to be adjusted; The safety distance between preset drones is dynamically adjusted through the safety distance adjustment formula. The safety distance adjustment formula is: ;in, Indicates the dynamically adjusted safe distance between drones; Indicates the power attenuation influence coefficient; Indicates the maximum battery power of the drone; Indicates the current battery level of the drone; Regard each UAV as a particle with repulsive characteristics, and regard the target points in the target monitoring area as source points with gravitational characteristics; when dij < dr, trigger the cooperative control mechanism of repulsive force and gravitational force, apply repulsive adjustment instructions according to the relative direction between UAVs, and guide the corresponding UAVs away from the collision path; and apply a guiding force in the direction of the target point to the UAV based on the position of the target point in the target monitoring area, guiding the UAV to fly towards the target monitoring area; superimpose the repulsive force and the guiding force to form the direction of the resultant force, and dynamically adjust the flight path of the corresponding UAV according to the direction of the resultant force; collect real-time air monitoring data, and the real-time air monitoring data includes measurement pollutant data, air temperature and humidity data, and wind speed and direction data in the target monitoring area.

4. The environmental assessment system based on real-time air monitoring data according to claim 3, characterized in that: The method for generating the continuously distributed pollutant concentration probability field includes: For the blind area of the UAV collection path, install a multi-band acoustic wave transmitting and receiving module on the UAV platform, and transmit multi-frequency acoustic wave signals in the direction of the blind area of the UAV collection path during the flight of the UAV; receive the echo signals after the acoustic wave diffracts at the edge of the obstacle and in the air medium; extract the characteristics of the echo signals, and the characteristics of the echo signals include the time delay of the echo signal, the intensity of the echo signal, and the phase change; For different pollutant concentrations and pollutant types, calculate the pollutant diffraction response function of multi-frequency acoustic waves in each frequency band, combine the pollutant diffraction response function and the characteristics of the echo signal, and use the least squares method for inversion to estimate the pollutant concentration in the blind area of the UAV collection path. Use the Kriging interpolation method to extend the inversion result to the entire target monitoring area to generate a continuously distributed pollutant concentration probability field.

5. The environmental assessment system based on real-time air monitoring data according to claim 4, characterized in that: The method for obtaining the pollutant concentration gradient includes; Divide the target monitoring area into multiple grid cells, and each grid cell corresponds to an estimated pollutant concentration; compare the pollutant concentration of any position point in each grid cell with the pollutant concentration of any position point in its surrounding adjacent grid cells, and extract the change amount of the pollutant concentration in the horizontal direction, vertical direction, and height direction of this position point, and combine the change amounts of the concentration in the horizontal direction, vertical direction, and height direction into a vector, representing the spatial gradient of the pollutant concentration at this position point; by extracting the change amount of the pollutant concentration for all grid cell position points, obtain the pollutant concentration gradient of the pollutant concentration probability field.

6. The environmental assessment system based on real-time air monitoring data according to claim 5, characterized in that: The method for generating the pollution distribution heat map includes: A pollutant concentration gradient threshold is preset. If the pollutant concentration gradient of any grid cell in the target monitoring area is greater than the preset pollutant concentration gradient threshold, it is determined that the grid cell needs to be encrypted and subdivided; if the pollutant concentration gradient of any grid cell in the target monitoring area is less than or equal to the preset pollutant concentration gradient threshold, it is determined that the grid cell does not need to be encrypted and subdivided; The grid cells that need to be refined and subdivided are divided into different sub-grid cells; the pollutant concentration is estimated again for each sub-grid cell; and the estimated pollutant concentrations of all grid cells are projected onto a map, using color depth to express different pollutant concentrations to obtain a pollution distribution heat map.

7. The environmental assessment system based on real-time air monitoring data according to claim 6, characterized in that: The method for obtaining the pollution source tracing data includes: The preset pollution emission characteristic database is composed of the emission characteristic vectors of the pollution sources, which include pollutant type data, pollutant emission intensity data and pollutant emission time data. The Lagrangian particle backtracking method is used to regard the pollutants in the pollution distribution heat map as particles, and the positions of their suspected source points are traced back. The particles are backtracked by the fourth-order Runge-Kutta method to obtain the entire particle trajectory. The pollutant type, pollutant concentration and pollution time at the starting point of the particle trajectory are extracted to form a pollutant characteristic vector. The pollutant characteristic vector is matched with the emission characteristic vector of the pollution source in the preset pollution emission characteristic database by cosine similarity to obtain the pollution source matching result, and the pollution source matching result is sorted in descending order according to the similarity to obtain the pollution tracing data.

8. The environmental assessment system based on real-time air monitoring data according to claim 7, characterized in that: The method for obtaining the environmental quality score includes: Construct and train an environmental quality assessment model, which includes an input layer, a GRU layer, a fully connected layer, and an output layer. The input layer of the environmental quality assessment model is used to input historical pollution source tracing data and population health data, and the output layer of the model is used to output the environmental quality score, using the identity function as the activation function. The environmental quality assessment model is a gated recurrent unit model, and the root mean square error is used as the loss function of the model. The current pollution source tracing data and population health data are input into the trained environmental quality assessment model to obtain the environmental quality score.

9. The environmental assessment system based on real-time air monitoring data according to claim 8, characterized in that: The method for determining whether the environmental quality of the area meets the standards includes: Preset the environmental quality score threshold, select any area, and compare the predicted environmental quality score with the preset environmental quality score threshold. If the predicted environmental quality score is greater than or equal to the preset environmental quality score threshold, the environmental quality of the area is judged to be up to standard; if the predicted environmental quality score is less than the preset environmental quality score threshold, the environmental quality of the area is judged to be not up to standard.

10. An environmental assessment method based on real-time air monitoring data, comprising: using the environmental assessment system based on real-time air monitoring data according to any one of claims 1 to 9, characterized in that: include: S1. Use nano gas-sensitive material sensor arrays to detect pollutants in the air and collect real-time wind field data; Combined with the drone swarm path planning algorithm, the drone collection path is dynamically adjusted under the guidance of real-time wind field data to collect real-time air monitoring data; S2. Compensate the air monitoring data within the blind spots of the drone's collection path using a multi-band acoustic diffraction compensation algorithm, thereby generating a continuously distributed pollutant concentration probability field. Calculate the spatial gradient of the pollutant concentration probability field to obtain the pollutant concentration gradient. S3. Using adaptive grid subdivision technology, we automatically refine regional grids based on air pollutant concentration gradients to generate pollution distribution heat maps. Based on these heat maps, we dynamically analyze the flow of pollutants using an air pollution source tracing algorithm. We also use a pre-set pollution emission characteristic database to trace pollution sources in different regions and generate pollution source tracing data. S4. Collect population health data, evaluate the environment of different regions based on pollution source tracing data and population health data, and predict environmental quality scores. For different regions, determine whether the environmental quality of the region meets the standards based on the predicted environmental quality scores; S5. If the environmental quality of the area meets the standard, the environmental quality of the area will be monitored every n periods of time and an environmental compliance report will be generated; if the environmental quality of the area does not meet the standard, an early warning report will be automatically generated and sent to the relevant management department.

Images