An unmanned aerial vehicle precision deodorization method and system based on odor concentration grading

Abstract

The application discloses an unmanned aerial vehicle precise deodorization method and system based on odor concentration grading, specifically, the odor concentration of different areas is sensed by an unmanned aerial vehicle in real time, and the areas are divided into different pollution grades, a horizontal grating scanning and multi-point vertical detection combined strategy is adopted for multi-point area scanning and data collection, an interpolation algorithm is introduced to construct a concentration field model, multi-section flight trajectories are planned, and the optimal deodorant type and concentration are further intelligently matched and executed, so that the high-altitude diffused odor can be precisely, efficiently and economically targeted and treated. The application can effectively solve the problem of single operation strategy, and improve the differentiated response emergency capability to pollution degrees.

Description

Technical Field

[0001] This invention relates to the field of drones and environmental governance, and more specifically, to a method and system for precise deodorization using drones based on odor concentration grading. Background Technology

[0002] In modern environmental governance, effectively controlling the unorganized emission of odors in open spaces is a common challenge, particularly evident in the following typical scenarios:

[0003] Firstly, modern intensive livestock farms typically employ enclosed environments and negative pressure ventilation systems. Large quantities of malodorous gases, including NH3 and H2S, generated inside these farms are collected by fans and discharged outside the farm via fixed elevated ventilation ducts or outlets. This results in high-concentration odors forming near the outlets, which dynamically change with wind direction and speed, spreading into an irregularly shaped, unevenly concentrated pollution plume in the atmosphere. This plume severely impacts air quality in the surrounding area, affecting residents' lives and potentially triggering social conflicts. Existing end-of-pipe treatment methods (such as in-duct spraying) only address the issue before emissions occur, becoming ineffective once the odor enters the atmosphere. Traditional ground-based fog cannons are also ineffective at effectively covering and controlling high-altitude pollution plumes, leading to poor emergency response results.

[0004] Secondly, odor problems at large urban landfills exhibit different characteristics. The odor sources are typically large, mobile, and unpredictable work surfaces, typical of "near-ground surface sources." Especially during the high temperatures of summer or after rain, the organic matter in the waste decomposes rapidly, producing large amounts of complex malodorous gases such as hydrogen sulfide, ammonia, and methanethiol. These gases diffuse unorganized from the vast work surfaces, forming even wider, more blurred-boundary, and more complex near-ground pollution plumes. Existing treatment methods, such as fixed sprinkler stations, have limited coverage; while ground-based mist cannons are ill-suited to the frequent changes in work surfaces and complex terrain, also exhibiting problems of blind spots and delayed response.

[0005] In summary, whether it's high-altitude point source pollution from livestock farms or near-ground non-point source pollution from landfills, existing remediation technologies face similar, deep-seated challenges: First, the sensing methods are outdated. Traditional methods cannot perform real-time three-dimensional morphology and internal concentration gradient surveys of dynamic, three-dimensional pollution plumes, resulting in a one-sided and delayed understanding of the pollution situation. Second, the operational strategies are simplistic. Although mobile monitoring and deodorization devices have emerged both domestically and internationally, their operational logic generally uses a single agent and fixed spraying parameters for indiscriminate operation, lacking the ability to respond differently to varying pollution levels. Therefore, how to establish a system that is universally applicable to various open pollution scenarios and can intelligently match and execute different aerial deodorization strategies based on real-time sensed odor concentration levels to achieve precise, efficient, and economical control of various pollution sources has become a pressing technical challenge in this field. Summary of the Invention

[0006] This invention overcomes the shortcomings of existing technologies and proposes a method and system for precise deodorization using drones based on odor concentration classification. Its purpose is to use drones to sense odor concentrations in different areas in real time and classify them into different pollution levels. For each pollution level, the system can intelligently match and execute the optimal type and concentration of deodorant, thereby achieving precise, efficient, and economical targeted treatment of odors diffused at high altitudes. This addresses the problems of current operational strategies being simplistic and lacking differentiated emergency response capabilities based on pollution levels.

[0007] The first aspect of this invention provides a method for precise deodorization of drones based on odor concentration grading, comprising:

[0008] S1: Based on the preset area, a map model is built. The drone carries out inspections according to the preset inspection route. Through the drone's sensor devices and GPS navigation module, it collects odor information, meteorological information and geographical information of the current location in real time and sends them to the ground control terminal.

[0009] S2: After receiving the information, the ground control terminal compares the odor information with the dynamic odor concentration trigger threshold. If the threshold is exceeded, the current location is determined to be the trigger point and the current task of the UAV is suspended.

[0010] S3: Using the trigger point as the center point, a local area is delineated in the map model, and a combination strategy of horizontal raster scanning and multi-point vertical detection is used in the local area to collect multi-point odor information and obtain an odor dataset.

[0011] S4: Using the three-dimensional kriging space interpolation algorithm, the odor dataset is interpolated in three-dimensional space to predict the concentration values ​​of three-dimensional grid points in the local area and obtain the concentration field model;

[0012] S5: Set the core area and optimal operating height according to the concentration field model, plan the flight trajectory through the path search algorithm, divide the concentration field model into layers based on the preset level standards, segment the flight trajectory according to the route of each layer, obtain multiple trajectories, and set segmented spraying instructions according to the multiple trajectories.

[0013] S6: Sends the flight trajectory and segmented spraying instructions to the drone for dynamic deodorization operations.

[0014] In this solution, S1 specifically refers to:

[0015] The sensor devices include odor monitoring sensors and weather sensors;

[0016] GPS navigation modules are used to collect geographic information in real time and transmit it to the ground control terminal. The geographic information includes flight altitude and three-dimensional coordinates.

[0017] Odor information includes dimensionless odor concentration values, and meteorological information includes wind direction and wind speed.

[0018] In this scheme, S2 includes:

[0019] Environmental meteorological data is acquired through ground control terminals. Based on the environmental meteorological data, combined with pollution standards and regional environmental sensitivity levels, odor concentration trigger thresholds are dynamically set.

[0020] Based on odor information, the odor concentration value is compared with the odor concentration trigger threshold. If the threshold is exceeded, the current position of the drone is determined to be the trigger point, and the drone is paused and enters a hovering standby state.

[0021] In this solution, S3 specifically refers to:

[0022] The local region is specifically a preset local boundary range, and its shape is a square.

[0023] In local areas, based on the current altitude of the drone, horizontal grating scanning is used to quickly collect odor information on the plane of the trigger point and identify one or more high-concentration odor areas in the horizontal direction.

[0024] Perform one or more vertical rise and fall detection and odor information collection in areas with high concentration of odor, and summarize the collected horizontal and vertical odor information to obtain a discrete odor dataset.

[0025] The odor dataset is then linked to the meteorological and geographic information of the data collection points.

[0026] In this solution, S4 specifically refers to:

[0027] In the 3D map, the odor dataset and the corresponding collection points are marked, and the odor dataset is weighted in 3D space using the 3D Kriging space interpolation algorithm.

[0028] The interpolation process involves determining the main diffusion direction under the current atmospheric conditions based on environmental meteorological data when calculating the correlation between any two points in the computation space. Points located in the downwind direction are assigned a high weight based on the main diffusion direction, while points located in the crosswind or vertical direction have a correlation inversely proportional to the distance value. Interpolation is then performed using correlation weighting to predict the odor concentration of the three-dimensional grid points in the entire local area and obtain the concentration field model.

[0029] In this solution, S5 specifically refers to:

[0030] Based on the concentration field model, areas with odor concentrations higher than the preset value are selected and designated as the core area. The optimal operating height is then set based on the vertical distribution curve of the core area.

[0031] All core areas are set as waypoints. Path search is performed in the map model based on the A* shortest path. During the search, for each path node, the search is based on the passage cost, which is inversely proportional to the odor concentration value, and a three-dimensional flight trajectory is obtained.

[0032] By pre-setting grade standards, the concentration field model is divided into layers to ensure that each layer corresponds to a grade standard;

[0033] The path segments belonging to different layers in the three-dimensional flight trajectory are divided into multiple trajectory segments;

[0034] In the multi-segment trajectory, the optimal deodorization command is matched by combining the dynamic dose response model to obtain the segmented spraying command. By controlling the metering pump and valve, different deodorant tanks are selected for spraying operations.

[0035] In this solution, S6 includes:

[0036] The flight trajectory and segmented spraying instructions are sent to the drone, which automatically executes the flight and spraying tasks, and executes corresponding agent switching and flow rate adjustment instructions according to the real-time location.

[0037] Save the data records of this operation and send them back to the ground control terminal to resume the execution of large-scale routine patrol and monitoring missions.

[0038] In this solution, the drone is equipped with deodorant tanks of various types and concentrations, metering pumps, valves, and atomizing nozzles.

[0039] A second aspect of the present invention also provides a drone-based precision deodorization system based on odor concentration classification. The system includes a memory, a processor, and a data interface. The memory includes a drone-based precision deodorization program based on odor concentration classification. When executed by the processor, the drone-based precision deodorization program based on odor concentration classification performs the following steps:

[0040] S1: Based on the preset area, a map model is built. The drone carries out inspections according to the preset inspection route. Through the drone's sensor devices and GPS navigation module, it collects odor information, meteorological information and geographical information of the current location in real time and sends them to the ground control terminal.

[0041] S2: After receiving the information, the ground control terminal compares the odor information with the dynamic odor concentration trigger threshold. If the threshold is exceeded, the current location is determined to be the trigger point and the current task of the UAV is suspended.

[0042] S3: Using the trigger point as the center point, a local area is delineated in the map model, and a combination strategy of horizontal raster scanning and multi-point vertical detection is used in the local area to collect multi-point odor information and obtain an odor dataset.

[0043] S4: Using the three-dimensional kriging space interpolation algorithm, the odor dataset is interpolated in three-dimensional space to predict the concentration values ​​of three-dimensional grid points in the local area and obtain the concentration field model;

[0044] S5: Set the core area and optimal operating height according to the concentration field model, plan the flight trajectory through the path search algorithm, divide the concentration field model into layers based on the preset level standards, segment the flight trajectory according to the route of each layer, obtain multiple trajectories, and set segmented spraying instructions according to the multiple trajectories.

[0045] S6: Sends the flight trajectory and segmented spraying instructions to the drone for dynamic deodorization operations.

[0046] A third aspect of the present invention also provides a computer-readable storage medium including a drone-based precision deodorization program based on odor concentration classification, wherein when the drone-based precision deodorization program based on odor concentration classification is executed by a processor, it implements the steps of the drone-based precision deodorization method based on odor concentration classification as described in any of the preceding claims.

[0047] The following beneficial effects can be achieved through the solution of the present invention:

[0048] This invention sets multi-level concentration thresholds and combines them with real-time survey data to objectively and quantitatively divide vague, continuous, and abstract odor pollution areas into multiple clear odor concentration level zones, thus solving the dilemma of traditional deodorization methods being unable to accurately locate the source of odor.

[0049] Traditional spraying methods typically employ a single concentration and uniform coverage approach, which faces the contradiction of mismatch between the dosage and concentration of the pesticide and the severity of the pollution area. The tiered spraying strategy of this invention, based on a three-dimensional concentration field model, can automatically and intelligently match the most suitable deodorant type and spray dosage for areas with different pollution levels. While ensuring treatment effectiveness, it greatly reduces ineffective pesticide consumption, significantly saves operating costs, and improves overall operational efficiency.

[0050] This invention utilizes spatial interpolation and path algorithms to perform dynamic flight planning and segmented spraying for unmanned aerial vehicles (UAVs), achieving intelligent and efficient three-dimensional spatial deodorization operations.

[0051] Traditional deodorization methods use corrosive or acidic agents, which, while effective in the short term, leave residues that cause secondary pollution to the soil and water sources of farms. This invention employs an environmentally friendly compound agent primarily composed of plant extracts and complex weak organic acids, combined with precision spraying technology, to eliminate secondary pollution at its source. This avoids the residual pollution to soil and water caused by traditional corrosive and acidic agents at the source; furthermore, by reducing the total amount of agent used, it further reduces the potential impact on the surrounding ecological environment, achieving green and sustainable odor control. Attached Figure Description

[0052] Figure 1 A flowchart of a drone-based precise deodorization method based on odor concentration classification according to the present invention is shown;

[0053] Figure 2 A structural diagram of the UAV of the present invention is shown;

[0054] Figure 3 A block diagram of a drone-based precision deodorization system based on odor concentration grading is shown. Detailed Implementation

[0055] To better understand the above-mentioned objects, features, and advantages of the present invention, the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments. It is worth noting that, unless otherwise specified, the embodiments and features described in these embodiments can be combined with each other.

[0056] Many specific details are set forth in the following description in order to provide a full understanding of the invention. However, the invention may also be practiced in other ways different from those described herein, and therefore the scope of protection of the invention is not limited to the specific embodiments disclosed below.

[0057] This invention relates to a method and system for precise odor removal by classifying and zoning odor pollution plumes formed by centralized emission outlets in open spaces such as farms using drones.

[0058] Figure 1 The flowchart of a drone-based precise deodorization method based on odor concentration classification according to the present invention is shown as a reference schematic diagram.

[0059] The first aspect of this invention provides a method for precise deodorization of drones based on odor concentration grading, comprising:

[0060] S1: Based on the preset area, a map model is built. The drone carries out inspections according to the preset inspection route. Through the drone's sensor devices and GPS navigation module, it collects odor information, meteorological information and geographical information of the current location in real time and sends them to the ground control terminal.

[0061] S2: After receiving the information, the ground control terminal compares the odor information with the dynamic odor concentration trigger threshold. If the threshold is exceeded, the current location is determined to be the trigger point and the current task of the UAV is suspended.

[0062] S3: Using the trigger point as the center point, a local area is delineated in the map model, and a combination strategy of horizontal raster scanning and multi-point vertical detection is used in the local area to collect multi-point odor information and obtain an odor dataset.

[0063] S4: Using the three-dimensional kriging space interpolation algorithm, the odor dataset is interpolated in three-dimensional space to predict the concentration values ​​of three-dimensional grid points in the local area and obtain the concentration field model;

[0064] S5: Set the core area and optimal operating height according to the concentration field model, plan the flight trajectory through the path search algorithm, divide the concentration field model into layers based on the preset level standards, segment the flight trajectory according to the route of each layer, obtain multiple trajectories, and set segmented spraying instructions according to the multiple trajectories.

[0065] S6: Sends the flight trajectory and segmented spraying instructions to the drone for dynamic deodorization operations.

[0066] According to an embodiment of the present invention, S1 specifically includes:

[0067] The sensor devices include odor monitoring sensors and weather sensors;

[0068] GPS navigation modules are used to collect geographic information in real time and transmit it to the ground control terminal. The geographic information includes flight altitude and three-dimensional coordinates.

[0069] Odor information includes dimensionless odor concentration values, and meteorological information includes wind direction and wind speed.

[0070] It's worth noting that the drone takes off and flies autonomously according to a pre-set inspection route. During flight, it collects real-time odor and meteorological information, such as dimensionless odor concentration, wind direction, and wind speed, using onboard odor monitoring and meteorological sensors. Simultaneously, the onboard GPS navigation module records the current flight location coordinates (specifically, three-dimensional coordinates, based on a map model to create a three-dimensional space) and flight altitude, sending these data to the ground control terminal. Upon receiving the data, the ground control terminal compares the real-time concentration value with a dynamic odor concentration trigger threshold.

[0071] According to an embodiment of the present invention, S2 includes:

[0072] Environmental meteorological data is acquired through ground control terminals. Based on the environmental meteorological data, combined with pollution standards and regional environmental sensitivity levels, odor concentration trigger thresholds are dynamically set.

[0073] Based on odor information, the odor concentration value is compared with the odor concentration trigger threshold. If the threshold is exceeded, the current position of the drone is determined to be the trigger point, and the drone is paused and enters a hovering standby state.

[0074] In a preferred embodiment, the trigger threshold is not a fixed value, but a concentration threshold dynamically set by the ground control terminal based on real-time acquired environmental meteorological data (such as temperature, humidity, wind speed, and wind direction), local environmental regulations and standards, and environmental sensitivity levels, to adapt to the odor diffusion patterns under different environmental conditions. Areas with higher environmental sensitivity and lower diffusion capacity can have a lower threshold set for determining the core area.

[0075] When the terminal determines that the received odor concentration value exceeds the current dynamic trigger threshold for the first time, it identifies the current location as a trigger point in an area of ​​excessive pollution. The system immediately records the three-dimensional coordinates of the trigger point and instructs the drone to pause its current patrol mission and enter a hovering standby state, ready to execute the next step.

[0076] The odor concentration value is the dimensionless odor concentration value.

[0077] According to an embodiment of the present invention, S3 specifically includes:

[0078] The local region is specifically a preset local boundary range, and its shape is a square.

[0079] In local areas, based on the current altitude of the drone, horizontal grating scanning is used to quickly collect odor information on the plane of the trigger point and identify one or more high-concentration odor areas in the horizontal direction.

[0080] Perform one or more vertical rise and fall detection and odor information collection in areas with high concentration of odor, and summarize the collected horizontal and vertical odor information to obtain a discrete odor dataset.

[0081] The odor dataset is then linked to the meteorological and geographic information of the data collection points.

[0082] In this invention, in order to accurately grasp the three-dimensional spatial distribution of the polluted area, the UAV takes the above-mentioned trigger point as the center and performs a three-dimensional spatial fine survey within a preset local boundary range.

[0083] During the data acquisition phase, a preset local boundary is defined centered on the aforementioned trigger point. This boundary is a square area with sides ranging from 30 to 100 meters. Within this boundary, a three-dimensional spatial survey route for detailed exploration is automatically planned. This survey route employs a combined strategy of "horizontal raster scanning + multi-point vertical detection": the UAV first performs a raster scan at its current altitude to quickly collect concentration data on the plane, thereby identifying one or more high-concentration hotspots in the horizontal direction. Subsequently, in the high-concentration hotspots, one or more vertical ascent and descent probes are performed to obtain detailed data on concentration changes in the vertical direction.

[0084] The ground control terminal receives all discrete data points (including 3D coordinates and odor concentration values) from the UAV survey and unifies them into the same geographic coordinate system, i.e., into the map model. At the same time, it binds the real-time environmental meteorological data (such as wind speed and wind direction) corresponding to the time of data collection for each data point, providing a data foundation for subsequent interpolation calculations.

[0085] According to an embodiment of the present invention, step S4 specifically includes:

[0086] In the 3D map, the odor dataset and the corresponding collection points are marked, and the odor dataset is weighted in 3D space using the 3D Kriging space interpolation algorithm.

[0087] The interpolation process involves determining the main diffusion direction under the current atmospheric conditions based on environmental meteorological data when calculating the correlation between any two points in the computation space. Points located in the downwind direction are assigned a high weight based on the main diffusion direction, while points located in the crosswind or vertical direction have a correlation inversely proportional to the distance value. Interpolation is then performed using correlation weighting to predict the odor concentration of the three-dimensional grid points in the entire local area and obtain the concentration field model.

[0088] In this embodiment, a three-dimensional digital concentration field model is generated using a spatial interpolation algorithm. After acquiring a preprocessed set of discrete data points, a three-dimensional kriging spatial interpolation algorithm based on atmospheric diffusion anisotropy is used. Unlike traditional kriging interpolation algorithms, which rely on physical distance when calculating the correlation between any two points in space, this algorithm introduces an effective distance dynamically corrected by real-time wind field data. First, the main diffusion direction under the current atmospheric conditions is determined based on wind speed and direction data. Then, when constructing the kriging interpolation algorithm, the distances in different directions are weighted: points located upwind, even if far away, are assigned a higher correlation weight because pollutants are transported further along the wind direction; conversely, the correlation of points located at or perpendicular to the wind direction decreases rapidly with increasing distance. Finally, through this interpolation algorithm, the system can calculate and predict the concentration values ​​of all three-dimensional grid points within the entire survey area. By incorporating environmental temperature, humidity, and wind speed, the interpolation results are made more consistent with fluid dynamics. Ultimately, a complete and intuitive three-dimensional digital concentration field model representing the pollution cloud in this local area is generated.

[0089] According to an embodiment of the present invention, S5 specifically includes:

[0090] Based on the concentration field model, areas with odor concentrations higher than the preset value are selected and designated as the core area. The optimal operating height is then set based on the vertical distribution curve of the core area.

[0091] All core areas are set as waypoints. Path search is performed in the map model based on the A* shortest path. During the search, for each path node, the search is based on the passage cost, which is inversely proportional to the odor concentration value, and a three-dimensional flight trajectory is obtained.

[0092] By pre-setting grade standards, the concentration field model is divided into layers to ensure that each layer corresponds to a grade standard;

[0093] The path segments belonging to different layers in the three-dimensional flight trajectory are divided into multiple trajectory segments;

[0094] In the multi-segment trajectory, the optimal deodorization command is matched by combining the dynamic dose response model to obtain the segmented spraying command. By controlling the metering pump and valve, different deodorant tanks are selected for spraying operations.

[0095] It is worth noting that the preset odor concentration is set according to the actual environment, and can also be selected as the area with the highest concentration. Each layer includes a certain spatial range.

[0096] The toll cost function is as follows:

[0097] ;

[0098] Where P is the passage cost of a certain node. The correction factor is D, the flight distance is T, the predicted time required to fly to a certain node is T, and the odor concentration value at a certain node is U (which can be a predicted value or a value that has already been collected).

[0099] It is worth noting that, in the intelligent operation planning based on the concentration field model, the path planning module of the ground control terminal performs in-depth analysis on the generated three-dimensional digital concentration field model and automatically plans the optimal operation route and spraying strategy.

[0100] By analyzing the concentration core area and determining the optimal operating height, the system identifies the region with the highest concentration value, known as the "concentration core area," through analysis of a 3D model. Simultaneously, by combining this with the vertical concentration distribution curve, it determines an "optimal operating height" that maximizes spraying efficiency and coverage.

[0101] This invention enables the planning of highly efficient three-dimensional flight trajectories. Based on the analyzed location of the concentration core area and the optimal operating altitude, the path planning module employs a path search algorithm to automatically plan a three-dimensional flight trajectory that efficiently traverses or circles the entire concentration core area. This trajectory ensures that the UAV can cover high-concentration areas to the maximum extent during operation.

[0102] The system spatially stratifies the entire three-dimensional concentration field model based on preset odor concentration impact levels (i.e., preset level standards, e.g., Level I: 0-20, Level II: 20-50, Level III: 50-70, Level IV: above 70). Then, the planned three-dimensional flight trajectory is segmented, with each segment corresponding to a specific concentration level region. Based on this, the system generates a series of segmented, differentiated spraying instructions. These instructions include: upon entering a certain concentration level region, the drone is instructed to automatically switch from its multiple (e.g., low, medium, and high) pre-loaded tanks containing different types or concentrations of deodorizer to the most suitable tank. Simultaneously, based on the precise concentration value at the current location, a dynamic dose response model calculates the optimal spray flow rate in real time and controls the metering pump and valves to execute precisely.

[0103] According to an embodiment of the present invention, S6 includes:

[0104] The flight trajectory and segmented spraying instructions are sent to the drone, which automatically executes the flight and spraying tasks, and executes corresponding agent switching and flow rate adjustment instructions according to the real-time location.

[0105] Save the data records of this operation and send them back to the ground control terminal to resume the execution of large-scale routine patrol and monitoring missions.

[0106] It's worth noting that during the precise deodorization operation and subsequent resumption of patrol, specifically, after receiving the task package containing a three-dimensional flight path and segmented spraying instructions, the drone automatically loads and strictly executes it. During the operation, the drone flies along the planned route and precisely executes corresponding agent switching and flow adjustment instructions based on its real-time location, completing a three-dimensional, graded, and targeted precise deodorization operation on the localized contaminated area. After the operation is completed, the system saves the data record of this operation. Subsequently, the drone completes its mission and resumes its large-scale routine patrol and monitoring tasks.

[0107] According to embodiments of the present invention, the drone is equipped with deodorant tanks of various types and concentrations, metering pumps, valves, and atomizing nozzles.

[0108] Figure 2 A structural diagram of the UAV of the present invention is shown.

[0109] According to an embodiment of the present invention, it further includes:

[0110] During the process of the drone executing segmented spraying instructions in each trajectory, a delay waiting time T1 is set, which is set according to the predicted flight time of each trajectory.

[0111] In each trajectory segment, odor information is collected after passing T1, and the odor value is fed back.

[0112] The original odor concentration value is obtained in the concentration field model, and the original odor concentration value is compared with the feedback odor value to obtain the instantaneous change value.

[0113] After trajectories of multiple segments, multiple instantaneous change values ​​are collected and serialized to obtain a feedback concentration change sequence;

[0114] The stationarity of the feedback concentration change sequence was verified by ADF stationarity check, and the moving average method was introduced to process the non-stationary part of the sequence.

[0115] Select one or more target path points from multiple path segments, and filter out corresponding data points from the feedback concentration change sequence based on the target path points;

[0116] By using a linear evaluation method, the linear rate of change of each data point in the feedback concentration change sequence is calculated, and the deodorization effectiveness of the target path point is evaluated by the linear rate of change.

[0117] Here, T1 is typically 1.5 to 2 times the predicted flight time, and this time period is used to collect the instantaneous change in odor information (odor concentration) after a certain delay following spraying. The predicted flight time is generally calculated using the current flight speed and the path length of the corresponding node. The instantaneous change value is the percentage increase or decrease in the feedback odor value compared to the original odor concentration value. One instantaneous change value corresponds to a trajectory segment or a path point. The target path point can be a core area point or an important spatial area point.

[0118] Traditional spray feedback analysis often lacks real-time dynamic spray analysis and effective evaluation processes. In drone deodorization evaluations under complex environments, the complex diffusion of air due to climate and environmental factors can mislead the feedback analysis, making it difficult to accurately evaluate deodorization feedback results in real time and in a timely manner. Therefore, this invention utilizes a concentration field model to analyze instantaneous odor concentration at each trajectory segment or path point, forming a feedback sequence based on the planned flight path. Through sequence stationarity analysis, nodes with anomalies are screened, and the sequence undergoes stationarization processing to effectively eliminate abnormal fluctuations and unexpected data points. This allows for the analysis of potential concentration change patterns under complex environmental conditions. Multi-point linear evaluation is then performed on the processed sequence, analyzing the corresponding linear change rate. The linear change rate is the rate of change at a certain point in the sequence; the smaller this value (closer to 0), the more stable the concentration change rate, indicating a more stable deodorization rate and higher deodorization effectiveness.

[0119] Example 1: Taking a large-scale intensive pig farm in City A as an example. Due to the low temperatures and strong winds in winter, which are unfavorable for the natural diffusion of odors, odor concentrations tend to accumulate in the downwind boundary area of ​​the farm, leading to complaints from nearby residents and necessitating emergency deodorization operations. The overall process can be... Figure 1 For reference only. The specific steps include the following:

[0120] Perform patrol monitoring and pollution identification:

[0121] First, after the drone system completes its self-check and passes, it takes off from the designated location. The operator pre-sets a cruise route along the entire perimeter of the farm via a ground control terminal, setting the cruise altitude to 15 meters and the cruise speed to 5 meters per second. The drone enters autonomous flight mode. During flight, odor monitoring sensors and weather sensors continuously collect data and transmit it back to the ground terminal in real time, along with GPS coordinates. The ground terminal's processor analyzes the transmitted concentration data in real time and compares it with a preset trigger threshold. In this embodiment, the threshold is set to a dimensionless odor concentration of 20 (corresponding to the starting value for a Level II minor impact), based on the upper limit of industry standards for certain areas of City A and the sensitivity of the on-site environment. When the system detects a concentration value exceeding 20 at a certain point, it immediately determines that an area of ​​excessive pollution has been discovered, automatically records the longitude, latitude, and altitude of the trigger point, and instructs the drone to suspend its cruise mission and hover near that point.

[0122] The specific dimensionless classification criteria for odor concentration are as follows:

[0123] Level I (0-20, no impact): Residents are usually unaware of the odor, it has no impact on human health, and does not affect livestock and poultry production;

[0124] Level II (20-50, slight impact): The odor is clearly audible, sensitive people may experience slight discomfort, livestock and poultry production may be damaged, and enterprises need to invest in basic deodorization costs.

[0125] Level III (50-70, significant impact): Residents' lives are severely disturbed by the odor, with most experiencing acute irritation symptoms; livestock and poultry production performance declines and diseases are prevalent; and enterprises face fines and rectification.

[0126] Level IV (70 or above, severe impact): Residents' normal lives are disrupted and there are health risks; livestock and poultry die on a large scale or lose their productive capacity; enterprises are severely punished or even shut down and go bankrupt.

[0127] Conduct detailed local surveys and construct a three-dimensional digital concentration field model:

[0128] The drone delineates a 50-meter-sided square area as the center of the trigger point for detailed local surveying. First, the drone performs a raster scan at its current altitude of 15 meters to quickly collect concentration data on the plane, identifying high-concentration hotspots in the horizontal direction. Then, above the identified high-concentration areas, it performs a vertical ascent and descent probe to acquire concentration distribution data in the vertical direction. After receiving all discrete survey data points, the ground terminal, based on the bound real-time wind speed and direction data, invokes a built-in three-dimensional kriging spatial interpolation algorithm based on atmospheric diffusion anisotropy to complete the data and build the model. Ultimately, it generates a three-dimensional digital concentration field model that visually represents the three-dimensional morphology, spatial location, and internal concentration gradient of the localized pollution area.

[0129] Intelligent operation planning based on concentration field model:

[0130] The path planning module of the ground terminal analyzes the generated 3D model. First, the concentration core area is determined using the traveling cube algorithm. Then, the optimal operating altitude is determined using the concentration-weighted centroid algorithm. Based on this, the system uses the A* search algorithm to automatically plan a 3D flight path for the UAV that can efficiently cover and traverse the concentration core area. Simultaneously, the system stratifies the entire concentration field model into levels I-IV (Level I: <20, Level II: 20-50, Level III: 50-70, Level IV: >70), matching differentiated spraying commands to different segments of the 3D flight path. For example, when the trajectory is in the Level II concentration area, the UAV is instructed to switch to tank 1 containing a low-concentration chemical deodorizer; when in the Level III area, it switches to tank 2 containing a medium-concentration agent; and when in the Level IV area, it switches to tank 3 containing a high-concentration agent. Furthermore, a dynamic dose response model is used to calculate and command the spraying flow rate in real time based on the specific concentration values ​​at the flight path points.

[0131] Perform precise deodorization and resume patrol:

[0132] After receiving and loading a mission package containing a complete 3D path and segmented spraying instructions, the drone begins automatic operation. The drone flies precisely along the planned route, and the central processor, based on its real-time location, sends instructions to the spraying subsystem to control the switching of the spray tanks and the spray flow rate, completing a three-dimensional, tiered, and targeted deodorization operation on the localized contaminated area. After the operation is completed, the operation log and related data are saved. Subsequently, the drone automatically returns to its cruise pause point to resume its large-scale cruise monitoring mission.

[0133] Example 2: This example illustrates the application scenario of a large landfill in City B during the summer. Summer's high temperature and humidity accelerate the fermentation of organic matter in the waste, leading to a large production and rapid volatilization of odorous gases such as hydrogen sulfide and ammonia. Simultaneously, the irregular changes in the landfill working surface make the odor source mobile and unpredictable. The objective of this example is to perform routine preventative inspections, proactively identifying and eliminating early, small-scale leaks exceeding safety standards before large-scale odor spread and user complaints arise. Specifically, this includes the following steps:

[0134] Implement cruise monitoring and odor monitoring thresholds:

[0135] In this embodiment, a routine inspection mode is adopted. The drone system automatically performs tasks daily during the period when odors are most volatile (10:00 AM to 4:00 PM). The drone, controlled by a ground terminal, sets a "bow"-shaped, carpet-like patrol route covering the entire landfill operation area and downwind sensitive areas. The patrol altitude is set at 10 meters to get closer to the ground source, and the patrol speed is 4 meters per second. In this embodiment, the threshold is based on the local standards of City B, the national standard GB 14554-93, and the characteristics of landfill management. Considering that humans are more sensitive to landfill odors, the following three-level dimensionless classification scheme is formulated, focusing on residents' lives, health, and enterprise management:

[0136] Level I (0-15, compliant and controllable): Residents have virtually no odor perception, no adverse health effects, and landfills only need to maintain routine sealing and drainage measures, with no risk of environmental complaints.

[0137] Level II (15-50, slightly exceeding the standard): The odor is clearly audible, and sensitive people may experience mild throat discomfort. The landfill needs to implement enhanced measures such as biological deodorization and faces complaints.

[0138] Level III (above 50, severely exceeding the standard): Residents' lives are severely affected; the landfill violates environmental regulations and faces the risk of fines, mandatory rectification, and closure.

[0139] Localized detailed surveys and construction of a three-dimensional concentration field model:

[0140] The drone first descends to a preset height of 5 meters above the ground, an altitude that effectively avoids ground obstacles while maximizing its proximity to the surface odor sources of the landfill. Then, at a low speed of 1 meter per second, the drone begins its east-west "grating-like" reciprocating flight along the north-south boundaries of the square work area. A fixed interval of 5 meters is maintained between each east-west flight path to ensure comprehensive coverage of the entire horizontal plane. During flight, onboard sensors continuously collect concentration data at a frequency of 5 times per second, transmitting it back to the ground terminal in real time via a wireless link.

[0141] After receiving and analyzing the data from the first phase, the ground terminal automatically calculates the center of the hotspot area. The drone then flies directly above this center point and hovers. Next, the drone performs a uniform vertical ascent, starting at a height of 3 meters and climbing at a speed of 0.5 meters per second to a height of 15 meters. Throughout the ascent, the sensors continuously collect concentration data at a high frequency, thus obtaining a complete concentration profile reflecting the vertical diffusion and elevation of pollutants. If multiple hotspots exist, the drone will perform a vertical detection on each hotspot in turn.

[0142] By employing an anisotropic 3D kriging interpolation algorithm and real-time wind field data (wind direction and speed) from an airborne weather station, this method accurately reproduces the physical phenomenon of odor diffusion under calm or turbulent wind conditions, primarily influenced by its own heat and molecular Brownian motion. This results in a high-fidelity 3D digital concentration field model, providing precise decision-making support for subsequent path planning. Under calm or turbulent conditions, the spatial correlation function attenuation coefficients of the algorithm tend to be consistent across all directions due to the lack of a stable dominant wind direction, leading to an interpolation result that more closely resembles an irregular "clump" model surrounding the source.

[0143] Intelligent job planning based on concentration field model:

[0144] Once the three-dimensional digital concentration field model is constructed, the intelligent operation planning module of the ground terminal immediately begins operation. Since the goal of this embodiment is to address early-stage, small-scale pollution, the core idea of ​​the planning is "rapid, thorough, and preventive of spread."

[0145] The improved 3D A* path planning algorithm demonstrates high adaptability in this scenario. Its internal pathfinding logic is specifically designed: when exploring possible paths from the starting point to the destination, the algorithm evaluates the "travel cost" of each node along the path. This cost is not simply the distance or time of flight, but is designed as a function inversely proportional to the odor concentration value of that node. This means that when faced with multiple choices, the algorithm will instinctively and preferentially choose the path that traverses areas with higher concentration values. As a result of this design, the algorithm ultimately generates not a simple straight line or shortest path, but a compact and meandering reciprocating flight trajectory that can efficiently traverse all areas with concentration values ​​exceeding 15 (i.e., Level II mild exceedance), thus achieving rapid and thorough eradication.

[0146] The odor composition of landfills is complex, and the optimal neutralizing agents vary depending on the composition. In this embodiment, the drone is pre-loaded with two different types of agents: the first agent tank contains a plant extract deodorizer. This agent's core components are citrus extract (whose main active ingredient is D-limonene), tea polyphenols, and various plant essential oils. Its mechanism of action is gentle, achieving deodorization by encapsulating, decomposing, and neutralizing odor molecules, without secondary pollution, making it an ideal choice for treating conventional or mildly excessive odors.

[0147] Box 2 contains a high-efficiency compound deodorizer. This agent is a formula specifically designed for high concentrations of malodors, and its components include: an alkaline component (potassium hydroxide) for rapidly neutralizing acidic gases (such as hydrogen sulfide); an oxidizing component (sodium percarbonate) for powerfully decomposing various organic odor molecules; and additives to enhance the agent's adhesion and adsorption effects.

[0148] When generating operation instructions, the system strictly follows these rules: for all areas along the flight path with concentration values ​​in the range of Level II (15-50), the system instructs the UAV to use tank 1 for spraying; and once the flight path enters an area with a concentration value of Level III (above 50), the system immediately instructs the UAV to switch to tank 2 for intensive treatment.

[0149] The spray flow rate is not a fixed value, but is precisely linearly correlated with the real-time concentration value to achieve the ultimate optimization of dosage.

[0150] The dynamic dose response model can be set up as follows:

[0151] Using the minimum concentration value of 15 that triggers spraying as a baseline, the corresponding spray flow rate is set to 5 ml / s. Above this baseline, the spray flow rate increases by 5 ml / s for every 10 units increase in odor concentration. A specific example of flow control based on this rule is as follows:

[0152] When the drone flew over an area with a real-time concentration of 25, the model calculated and instructed the metering pump to output a flow rate of 10 ml / s.

[0153] When the drone enters an area where the concentration of 65 is severely exceeded, the system will switch the command to the No. 2 medicine tank and calculate the corresponding flow rate of 30 ml / s according to the same linear rule to ensure that the potent medicine is sprayed in sufficient dose to deal with the severe pollution.

[0154] Precise deodorization operation and resumption of inspection:

[0155] After loading the mission package, the drone automatically executes the operation. Following a planned, compact flight path, it efficiently and comprehensively targets and sprays the entire early-stage contamination plume. Upon completion, the system logs the event (including location, time, concentration, main pollutant components, type and dosage of reagents used), providing valuable data support for subsequent analysis of landfill operation management and optimization. The drone then automatically returns to its cruise pause point to continue its routine, comprehensive inspection mission.

[0156] The concentration values ​​shown in the examples are relative reference standard values. They are based on different types of odors and deodorization conditions in different scenarios. The numerical standardization range and threshold can be adjusted to adapt to different odor environments.

[0157] Figure 3 A block diagram of a drone-based precision deodorization system based on odor concentration grading is shown.

[0158] A second aspect of the present invention also provides a drone-based precision deodorization system based on odor concentration grading. This system includes: a memory, a processor, and a data interface. The data interface is used to connect to devices such as a drone, a ground control terminal, and a user mobile terminal. The memory stores a series of data collected and analyzed by the system during the deodorization process. The memory also includes a drone-based precision deodorization program based on odor concentration grading. When executed by the processor, the drone-based precision deodorization program based on odor concentration grading performs the following steps:

[0159] S1: Based on the preset area, a map model is built. The drone carries out inspections according to the preset inspection route. Through the drone's sensor devices and GPS navigation module, it collects odor information, meteorological information and geographical information of the current location in real time and sends them to the ground control terminal.

[0160] S2: After receiving the information, the ground control terminal compares the odor information with the dynamic odor concentration trigger threshold. If the threshold is exceeded, the current location is determined to be the trigger point and the current task of the UAV is suspended.

[0161] S3: Using the trigger point as the center point, a local area is delineated in the map model, and a combination strategy of horizontal raster scanning and multi-point vertical detection is used in the local area to collect multi-point odor information and obtain an odor dataset.

[0162] S4: Using the three-dimensional kriging space interpolation algorithm, the odor dataset is interpolated in three-dimensional space to predict the concentration values ​​of three-dimensional grid points in the local area and obtain the concentration field model;

[0163] S5: Set the core area and optimal operating height according to the concentration field model, plan the flight trajectory through the path search algorithm, divide the concentration field model into layers based on the preset level standards, segment the flight trajectory according to the route of each layer, obtain multiple trajectories, and set segmented spraying instructions according to the multiple trajectories.

[0164] S6: Sends the flight trajectory and segmented spraying instructions to the drone for dynamic deodorization operations.

[0165] When the system is running, it can perform one or more steps of the above-described method for precise deodorization of drones based on odor concentration classification.

[0166] A third aspect of the present invention also provides a computer-readable storage medium including a drone-based precision deodorization program based on odor concentration classification, wherein when the drone-based precision deodorization program based on odor concentration classification is executed by a processor, it implements the steps of the drone-based precision deodorization method based on odor concentration classification as described in any of the preceding claims.

[0167] This invention discloses a method and system for precise odor removal using drones based on odor concentration classification. Specifically, the drone senses the odor concentration in different areas in real time and classifies them into different pollution levels. A combined strategy of horizontal grating scanning and multi-point vertical detection is employed for multi-point area scanning and data acquisition. An interpolation algorithm is introduced to construct a concentration field model, plan multiple flight trajectories, and further intelligently match and execute the optimal type and concentration of deodorant. This achieves precise, efficient, and economical targeted treatment of odors diffused at high altitudes. This invention effectively solves the problem of current single-strategy operations and improves the emergency response capability to differentiated pollution levels.

[0168] In the above embodiments, implementation can be achieved, in whole or in part, through software, hardware, firmware, or any combination thereof. When implemented using software, it can be implemented, in whole or in part, as a computer program product. The computer program product includes one or more computer instructions. When the computer instructions are loaded and executed on a computer, the processes or functions described in the embodiments of this application can be generated, in whole or in part. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired (e.g., coaxial cable, fiber optic cable, data subscriber line) or wireless (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium can be any available medium accessible to a computer or a data storage device such as a server or data center that integrates one or more available media. The available medium may be a magnetic medium (e.g., floppy disk, hard disk, magnetic tape), an optical medium (e.g., digital universal optical disc), or a semiconductor medium (e.g., solid-state drive). In the various embodiments of this application, unless otherwise specified or logically conflicting, the terminology and / or descriptions between different embodiments are consistent and can be mutually referenced. Technical features in different embodiments can be combined to form new embodiments based on their inherent logical relationships.

[0169] In the embodiments of this application, "at least one" refers to one or more, and "more than one" refers to two or more. "And / or" describes the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A existing alone, A and B existing simultaneously, or B existing alone, where A and B can be singular or plural. In the textual description of the embodiments of this application, the character " / " generally indicates that the preceding and following related objects have an "or" relationship. In this application, "first," "second," and various numerical designations are only for the convenience of description and are not used to limit the scope of the embodiments of this application. For example, they are used to distinguish different messages, rather than to describe a specific order or sequence.

[0170] It is understood that the various numerical designations used in the embodiments of this application are merely for descriptive convenience and are not intended to limit the scope of the embodiments of this application. The order of the process numbers does not imply the order of execution; the execution order of each process should be determined by its function and internal logic.

[0171] Finally, it should be noted that the above description is only a specific implementation of this application, but the protection scope of this application is not limited thereto. Any changes or substitutions within the technical scope disclosed in this application should be covered within the protection scope of this application.

Claims

1. A method for precise odor removal using drones based on odor concentration grading, characterized in that, include: S1: Based on the preset area, a map model is built. The drone carries out inspections according to the preset inspection route. Through the drone's sensor devices and GPS navigation module, it collects odor information, meteorological information and geographical information of the current location in real time and sends them to the ground control terminal. S2: After receiving the information, the ground control terminal compares the odor information with the dynamic odor concentration trigger threshold. If the threshold is exceeded, the current location is determined to be the trigger point and the current task of the UAV is suspended. S3: Using the trigger point as the center point, a local area is delineated in the map model, and a combination strategy of horizontal raster scanning and multi-point vertical detection is used in the local area to collect multi-point odor information and obtain an odor dataset. S4: Using the three-dimensional kriging space interpolation algorithm, the odor dataset is interpolated in three-dimensional space to predict the concentration values ​​of three-dimensional grid points in the local area and obtain the concentration field model; S5: Set the core area and optimal operating height according to the concentration field model, plan the flight trajectory through the path search algorithm, divide the concentration field model into layers based on the preset level standard, segment the flight trajectory according to the route of each layer, obtain multiple trajectories, and set segmented spraying instructions according to the multiple trajectories. S6: Sends flight trajectory and segmented spraying instructions to the drone for dynamic deodorization operations; Specifically, S3 is: The local region is specifically a preset local boundary range, and its shape is a square. In local areas, based on the current altitude of the drone, horizontal grating scanning is used to quickly collect odor information on the plane of the trigger point and identify one or more high-concentration odor areas in the horizontal direction; Perform one or more vertical rise and fall detection and odor information collection in areas with high concentration of odor, and summarize the collected horizontal and vertical odor information to obtain a discrete odor dataset. The odor dataset is then linked to the meteorological and geographic information of the data collection points. Specifically, S4 is: In the 3D map, the odor dataset and the corresponding collection points are marked, and the odor dataset is weighted in 3D space using the 3D Kriging space interpolation algorithm. The interpolation process involves determining the main diffusion direction under the current atmospheric environment based on environmental meteorological data when calculating the correlation between any two points in the computation space. Based on the main diffusion direction, points located in the downwind direction are assigned a high weight, while points located in the crosswind or vertical direction have a correlation inversely proportional to the distance value. Interpolation is then performed using correlation weighting to predict the odor concentration of the three-dimensional grid points in the entire local area and obtain the concentration field model. Specifically, S5 is: Based on the concentration field model, areas with odor concentrations higher than the preset value are selected and designated as the core area. The optimal operating height is then set based on the vertical distribution curve of the core area. All core areas are set as waypoints. Path search is performed in the map model based on the A* shortest path. During the search, for each path node, the search is based on the passage cost, which is inversely proportional to the odor concentration value, and a three-dimensional flight trajectory is obtained. By pre-setting grade standards, the concentration field model is divided into layers to ensure that each layer corresponds to a grade standard; The path segments belonging to different layers in the three-dimensional flight trajectory are divided into multiple trajectory segments; In the multi-segment trajectory, the optimal deodorization command is matched by combining the dynamic dose response model to obtain the segmented spraying command. By controlling the metering pump and valve, different deodorant tanks are selected for spraying operations.

2. The method for precise deodorization of drones based on odor concentration classification according to claim 1, characterized in that, Specifically, S1 is: The sensor devices include odor monitoring sensors and weather sensors; The GPS navigation module is used to collect geographic information in real time and transmit it to the ground control terminal. The geographic information includes flight altitude and three-dimensional coordinates. Odor information includes dimensionless odor concentration values, and meteorological information includes wind direction and wind speed.

3. The method for precise deodorization of drones based on odor concentration classification according to claim 1, characterized in that, S2 includes: Environmental meteorological data is acquired through ground control terminals. Based on the environmental meteorological data, combined with pollution standards and regional environmental sensitivity levels, odor concentration trigger thresholds are dynamically set. Based on odor information, the odor concentration value is compared with the odor concentration trigger threshold. If the threshold is exceeded, the current position of the drone is determined to be the trigger point, and the current task of the drone is paused and enters a hovering standby state.

4. The method for precise deodorization of drones based on odor concentration classification according to claim 1, characterized in that, S6 includes: The flight trajectory and segmented spraying instructions are sent to the drone, which automatically executes the flight and spraying tasks, and executes corresponding agent switching and flow rate adjustment instructions according to the real-time location. Save the data records of this operation and send them back to the ground control terminal to resume the execution of large-scale routine patrol and monitoring missions.

5. The method for precise deodorization of drones based on odor concentration classification according to claim 1, characterized in that, The drone is equipped with deodorant tanks of various types and concentrations, metering pumps, valves, and atomizing nozzles.

6. A drone-based precision deodorization system based on odor concentration grading, characterized in that, The system includes: a memory, a processor, and a data interface. The memory includes a drone-based precision deodorization program based on odor concentration classification. When the processor executes the drone-based precision deodorization program based on odor concentration classification, it implements the steps of the drone-based precision deodorization method as described in claim 1.

7. A computer-readable storage medium, characterized in that, The computer-readable storage medium includes a drone-based precision deodorization program based on odor concentration classification. When the drone-based precision deodorization program based on odor concentration classification is executed by a processor, it implements the steps of the drone-based precision deodorization method based on odor concentration classification as described in any one of claims 1 to 5.

Images