Indoor air quality inspection robot

Abstract

The invention discloses an indoor air quality inspection robot, which belongs to the field of indoor air quality inspection, and comprises the following steps: collecting air quality data in real time by adopting a sensor network according to preliminary monitoring priority distribution, extracting abnormal fluctuation characteristics from air quality change of each region, and determining a target region range needing key inspection; a path planning algorithm is called to generate a shortest feasible path from a current position to a target area through a local path adjustment scheme in combination with a global path optimization target, and a real-time updating result of a dynamic inspection path is obtained; and transmitting task allocation data of each inspection robot through a scheduling instruction of multi-device collaborative operation, and extracting target area and path planning information of each device from the task allocation data to obtain a coordination optimization scheme of overall inspection coverage.

Description

Technical Field

[0001] This invention belongs to the field of indoor air quality inspection, and in particular relates to an indoor air quality inspection robot. Background Technology

[0002] Indoor air quality monitoring and improvement is a crucial area in modern smart home and public space management, directly impacting people's health and quality of life. With accelerating urbanization and increasing complexity of indoor environments, efficient and accurate air quality monitoring has become an issue that cannot be ignored. Especially in large indoor spaces, where air quality distribution is uneven and potential pollution sources are complex and diverse, traditional monitoring methods often fall short of meeting the demands for real-time and comprehensive monitoring.

[0003] While some current solutions achieve air quality detection to a certain extent, they generally suffer from insufficient dynamic adaptability. Many methods rely on sensors at fixed locations or inspection equipment along pre-defined paths, making it difficult to cope with sudden changes in the indoor environment, such as the appearance of temporary obstacles or the dynamic migration of air pollution sources. This static design approach leads to incomplete monitoring coverage and limits the timeliness and accuracy of data collection, especially in complex scenarios involving multiple areas and multiple tasks, where coordination between devices and flexibility in path planning are particularly lacking.

[0004] Therefore, how to achieve efficient coverage of all key areas by inspection robots in complex indoor environments through dynamic path planning, while taking into account real-time data transmission and multi-device collaborative operation, has become a key problem that this research urgently needs to solve. Summary of the Invention

[0005] To address the aforementioned technical problems, this invention provides an indoor air quality inspection robot, comprising:

[0006] The environmental perception module is used to acquire information on the indoor space layout and obstacle distribution;

[0007] The area division module is used to divide the inspection area and determine the monitoring priority based on the indoor space layout and obstacle distribution information.

[0008] The data acquisition module is used to collect air quality data in the inspection area and extract abnormal fluctuation characteristics to obtain the range of key inspection target areas;

[0009] The dynamic monitoring module is used to acquire the current location and obstacle dynamic information based on the scope of the key inspection target area, and adjust the local path accordingly;

[0010] The path planning module is used to generate a dynamic inspection path based on the local path;

[0011] The resource scheduling module is used to move and collect data based on the dynamic inspection path to determine the location of the pollution source;

[0012] The priority adjustment module is used to adjust the priority of monitoring points and allocate inspection resources based on the location of the pollution source, and generate inspection instructions.

[0013] The collaborative optimization module is used to optimize the overall inspection coverage coordination based on the inspection instructions and generate a coordination optimization scheme.

[0014] The coverage detection module is used to update the inspection path and tasks based on the coordination optimization scheme and trigger supplementary inspection tasks.

[0015] The task generation module is used to generate a coverage path and determine an inspection execution plan based on the supplementary inspection task.

[0016] Preferably, the environmental perception module extracts spatial layout and obstacle distribution data from the indoor environment through a pre-constructed three-dimensional model, generates an initial inspection area division map, and uses data extraction technology to classify the spatial layout and obstacle distribution data, mark high-complexity areas, and determine the monitoring priority of high-complexity areas.

[0017] Preferably, the data acquisition module collects data from each region through a sensor network to obtain the raw data set of air quality, and uses real-time analysis technology to continuously track the air quality change trend of each region, extract the change characteristics in the time series, and determine potential abnormal fluctuation ranges.

[0018] Preferably, the dynamic monitoring module obtains the current location information and surrounding obstacle distribution data of the target area from the sensors carried by the inspection robot through a real-time data transmission mechanism, and uses a pre-established obstacle recognition model to analyze whether there are characteristics of temporary obstacles in the obstacle distribution data, and determines whether there is a risk of obstruction.

[0019] Preferably, the path planning module acquires environmental data and obstacle information from the current location, combines it with the coordinate data of the target area, constructs an initial path planning model, integrates real-time environmental change data of the local path, and uses an adjustment strategy to dynamically correct the path to determine the feasibility of the local path.

[0020] Preferably, the resource scheduling module drives the inspection robot to move along the designated path according to the real-time update results of the dynamic inspection path, while collecting air quality data along the way, extracting significant patterns of pollution distribution using feature analysis methods, and classifying the data using the support vector machine algorithm to determine the distribution characteristics of potential pollution areas.

[0021] Preferably, the priority adjustment module obtains the location information and characteristic values ​​of pollution sources from the environmental monitoring system, constructs a dynamic distribution map of regional divisions, and compares the characteristic values ​​of each regional division with preset thresholds according to the pollution distribution status, marks high-priority areas, and determines the monitoring scope that needs to be focused on.

[0022] Preferably, the collaborative optimization module obtains the task allocation number of the inspection robot from the data transmission chain through the multi-device collaborative scheduling instruction set, determines the initial task distribution of each device, and generates the path planning map of each inspection robot by combining the target area point and the area division method, thereby obtaining the preliminary inspection range allocation.

[0023] Preferably, the coverage detection module acquires the real-time location data and current path planning information of the inspection robot, compares the data with the preset area division information, determines whether there are uncovered areas, and dynamically adjusts the path planning of the inspection robot according to the range and priority of the uncovered areas.

[0024] Preferably, the task generation module acquires the location data of the uncovered area, uses a geographic information system to scan and analyze the target area, identifies the area not covered by the inspection, generates a coverage path through a path planning algorithm, and uses Dijkstra's algorithm to calculate the shortest path from the current inspection point to the uncovered area, thus obtaining optimized coverage path data.

[0025] Compared with the prior art, the present invention has the following advantages and technical effects:

[0026] This invention discloses a dynamic indoor environmental inspection method. It acquires the indoor layout and obstacle distribution through a pre-established 3D model, constructs an initial inspection area division map, and determines monitoring priorities. A sensor network is used to collect air quality data in real time, extracting abnormal fluctuation characteristics to identify key inspection areas. Combined with real-time obstacle information, the inspection path is dynamically adjusted, driving a robot to move along the optimal path and collect data. Pollution source distribution characteristics are extracted from the collected data to determine the location of potential pollution sources and adjust monitoring priorities. Through multi-device collaborative operation, the overall inspection coverage scheme is optimized, and the paths and tasks of each robot are updated in real time. Supplementary inspection tasks are triggered for uncovered areas to ensure comprehensive coverage. This invention achieves intelligent dynamic inspection in complex indoor environments, improving inspection efficiency and the accuracy of pollution source location. Attached Figure Description

[0027] The accompanying drawings, which form part of this application, are used to provide a further understanding of this application. The illustrative embodiments and descriptions of this application are used to explain this application and do not constitute an undue limitation of this application. In the drawings:

[0028] Figure 1 This is a schematic diagram of the system structure according to an embodiment of the present invention. Detailed Implementation

[0029] It should be noted that, unless otherwise specified, the embodiments and features described in this application can be combined with each other. This application will now be described in detail with reference to the accompanying drawings and embodiments.

[0030] It should be noted that the steps shown in the flowchart in the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions, and although a logical order is shown in the flowchart, in some cases the steps shown or described may be executed in a different order than that shown here.

[0031] Example 1

[0032] like Figure 1 As shown, this embodiment provides an indoor air quality inspection robot, including:

[0033] The environmental perception module is used to acquire information on the indoor space layout and obstacle distribution;

[0034] By using a pre-established 3D model of the indoor environment, the layout data of the indoor space and the distribution information of obstacles are obtained. Based on the dynamic characteristics of the complex indoor environment, an initial inspection area division map is constructed to obtain the preliminary monitoring priority distribution of each area.

[0035] Using a pre-constructed 3D model, spatial layout and obstacle distribution data are extracted from the indoor environment to generate an initial inspection area division map, obtaining preliminary distribution information for each area. Data extraction techniques are employed to classify the spatial layout and obstacle distribution data, determining the complexity of each area. If the obstacle distribution density in a certain area exceeds a preset threshold, it is marked as a high-complexity area, and the monitoring priority for high-complexity areas is determined.

[0036] Specifically, using a pre-established 3D model of the indoor environment, the layout data and obstacle distribution information of the indoor space are first obtained using laser scanning technology and point cloud data processing algorithms. The specific method is to use a LiDAR device to scan the indoor environment, generate point cloud data, and then use Euclidean clustering algorithm to segment the point cloud into different object regions. Assuming the total indoor area is 500 square meters, fixed obstacles such as walls and furniture are identified, occupying 200 square meters, and the remaining 300 square meters are passable areas. The analysis shows that the obstacle distribution density is 0.4. The layout data is stored in the form of a grid, with each grid cell being 0.5 meters × 0.5 meters, for subsequent calculations. Next, considering the dynamic characteristics of the complex indoor environment, an initial inspection area division map was constructed. A Voronoi diagram-based area segmentation algorithm was used to divide the indoor space into 10 sub-regions, each with an area between 30 and 50 square meters. By analyzing obstacle density and historical dynamic change data (such as personnel flow frequency, assuming an average daily flow of 50 times in a certain area) within each region, the dynamic risk coefficient for each region was calculated. The risk coefficient formula is R = 0.6 × density + 0.4 × flow frequency, resulting in an R value of 0.8 for a high-risk area. Finally, based on the risk coefficients, a preliminary monitoring priority distribution for each region was obtained. The priority calculation used a weighted average method, combining the risk coefficient R and the area proportion (assuming a certain area's proportion is 0.1), with the priority P = 0.7 × R + 0.3 × area proportion. The calculation results show that the priority for a high-risk area is 0.59.

[0037] The area division module is used to divide the inspection area and determine the monitoring priority based on the indoor space layout and obstacle distribution information.

[0038] Based on the preliminary monitoring priority distribution, a sensor network is used to collect air quality data in real time, extract abnormal fluctuation characteristics from the air quality changes in each region, and determine the target area range that needs to be inspected.

[0039] Data is collected from various regions using a sensor network to obtain raw air quality datasets, which are stored in a pre-established database to obtain preliminary air quality distribution information. Based on this preliminary information, real-time analysis technology is used to continuously track air quality trends in each region, extracting change features from the time series to identify potential abnormal fluctuation ranges. For potential abnormal fluctuation ranges, if the change features exceed a preset threshold, feature extraction methods are used to quantify the abnormal fluctuations, obtaining specific fluctuation feature descriptions. Based on these descriptions, combined with monitoring priorities and regional distribution information, high-priority target areas are identified to determine which areas exhibit significant air quality anomalies.

[0040] Specifically, in the process of determining the target areas for air quality monitoring and inspection, air quality data for each area is first collected in real time through a sensor network. Assuming 100 sensor nodes are deployed, covering 10 major urban areas, PM2.5, PM10, and sulfur dioxide data are collected hourly. The data is stored in a cloud database in micrograms per cubic meter. Each node generates approximately 0.5 MB of data per hour, for a total of 50 MB. Noise is removed using data cleaning algorithms, such as median filtering to smooth PM2.5 readings and removing outliers exceeding three standard deviations to ensure data accuracy. Next, based on the initial monitoring priority distribution, resources are densely allocated to high-pollution areas. For example, historical data shows that the annual average PM2.5 value for areas A and B exceeds 50 micrograms per cubic meter, so these areas are prioritized as Level 1, while other areas are prioritized as Level 2 or 3. The system automatically adjusts the sensor sampling frequency: Level 1 areas are sampled every 30 minutes, and Level 2 areas are sampled hourly to increase data density. Subsequently, abnormal fluctuation characteristics are extracted from the air quality changes in various regions. Using time series analysis algorithms, such as the sliding window method with a 24-hour window, the standard deviation of PM2.5 data is calculated. If the standard deviation of a region exceeds a preset threshold of 15 micrograms per cubic meter, it is marked as an abnormal fluctuation. Simultaneously, an autoregressive integral moving average (ARIMA) model is used to predict the trend over the next 6 hours. If the predicted value continues to rise and exceeds the national standard of 35 micrograms per cubic meter, the anomaly is further confirmed. Finally, the target areas requiring focused inspection are determined. Based on the abnormal fluctuation characteristics and prediction results, the system calculates priorities using the comprehensive pollution index (API). The API formula is a weighted average of the concentrations of each pollutant and the standard value. If the API value of region A is 120, exceeding the threshold of 100, while region B is 90, region A is automatically listed as a key inspection target, and an inspection task sheet is generated, including specific latitude and longitude coordinates and pollution data details.

[0041] The data acquisition module is used to collect air quality data in the inspection area and extract abnormal fluctuation characteristics to obtain the range of key inspection target areas;

[0042] For the target area, a real-time data transmission mechanism is used to obtain dynamic updates on the current location of the inspection robot and the distribution of surrounding obstacles, and to determine whether there are temporary obstacles blocking the way. If there are temporary obstacles, the local path adjustment plan is recalculated.

[0043] Through a real-time data transmission mechanism, the current location information and surrounding obstacle distribution data within the target area are acquired from the sensors mounted on the inspection robot, yielding preliminary environmental perception results. Based on these preliminary results, a pre-established obstacle recognition model is used to analyze the characteristics of temporary obstacles in the obstacle distribution data to determine if there is a risk of obstruction. If a risk of obstruction by temporary obstacles is identified, key points are extracted from the current location information and obstacle distribution data to determine the affected local path range.

[0044] Specifically, to address the dynamic path adjustment requirements of the inspection robot within the target area, a real-time data transmission mechanism is first used to update the current position data every second using the GPS module and LiDAR sensor mounted on the robot. Assuming the current coordinates are (X=25.5, Y=18.3) meters, the LiDAR scans the obstacle distribution within a 5-meter radius, acquires point cloud data, and generates a two-dimensional grid map with a resolution of 0.1 meters. An object with a width of 1.2 meters and a height of 0.8 meters is detected 2.3 meters ahead. Based on historical data analysis, this object is not in the fixed obstacle database and is determined to be a temporary obstacle. Next, the system automatically invokes a path planning algorithm, such as the A* algorithm. Inputting the current grid map and the target point coordinates (X = 30.0, Y = 20.0 meters), the algorithm calculates the optimal path. Starting from the current point, it evaluates the movement costs in eight surrounding directions, prioritizing the path with the lowest total cost, avoiding temporary obstacle areas. The adjusted local path is to detour to (X = 26.8, Y = 19.1 meters) before returning to the main path. The path length increases by 1.5 meters, and the estimated time increases by 10 seconds. The system updates the path information to the robot control module in real time to ensure it follows the new path.

[0045] The dynamic monitoring module is used to acquire the current location and obstacle dynamic information based on the scope of the key inspection target area, and adjust the local path accordingly;

[0046] By combining local path adjustment schemes with global path optimization objectives, a path planning algorithm is invoked to generate the shortest feasible path from the current location to the target area, thus obtaining the real-time update results of the dynamic inspection path.

[0047] By acquiring environmental data and obstacle information from the current location and combining it with the coordinate data of the target area, an initial path planning model is constructed, resulting in a preliminary path framework. Based on this preliminary path framework, real-time environmental change data of local paths is integrated, and an adjustment strategy is used to dynamically correct the path, determining the feasibility of local paths. For the feasibility of local paths, combined with the optimization objective of the global path, a path planning algorithm is invoked to calculate the overall path, obtaining candidate shortest path solutions. Using the candidate shortest path solutions, the real-time requirements of dynamic inspection are analyzed. If environmental obstacles or path blockages are detected, the local path is recalculated, resulting in an updated path segment. Based on the updated path segment, the constraints of feasible paths are integrated to determine whether the path meets the time and safety requirements of the inspection task, thus determining the final feasible path solution.

[0048] Specifically, in the process of realizing real-time updates of dynamic inspection paths, the current position is first analyzed in real time through a local path adjustment scheme. Assuming the current position coordinates are (10.5, 20.3) and the center point of the target area is (50.8, 60.2), the system uses the A* algorithm to perform preliminary path search and calculates feasible paths to avoid obstacles within the local area. The obstacle data comes from real-time sensor scanning, and the obstacle positions are (20.1, 30.4) and (30.7, 40.9). The algorithm calculates the local shortest path length as 15.7 meters using a gridded map (grid size of 1.0 meter) in 0.02 seconds, ensuring no collisions on the local path. Next, combining the global path optimization objective, the system integrates the local path with the global objective. Dijkstra's algorithm is used to calculate the shortest path from the current location to the target area. Considering known fixed obstacles and dynamic traffic flow data in the global map, the system assumes a total global path length of 45.3 meters, containing three key nodes: (15.2, 25.6), (30.5, 40.8), and (45.1, 55.3). After optimization, the path smoothness is improved by 12%, and the time cost is reduced by 8%. By analyzing the path curvature and the distance between nodes, the path is ensured to be suitable for the inspection robot to travel at a speed of 2.5 meters per second. Subsequently, the path planning algorithm is invoked to generate the final path. Combining the local and global results, the path is smoothed using Bézier curves, generating continuous path points with a point spacing of 0.5 meters and a total of 90 points to ensure path continuity.

[0049] The path planning module is used to generate a dynamic inspection path based on the local path;

[0050] Based on the real-time updates of the dynamic inspection path, the inspection robot is driven to move along the designated path, while collecting air quality data along the way. The possible distribution characteristics of pollution sources are extracted from the collected data to determine the location information of potential temporary pollution sources.

[0051] A dynamic inspection mechanism is used to acquire the real-time movement status of the inspection robot along a designated route. Combined with path update information, the movement drive commands are adjusted to obtain location data during the inspection process. Based on the location data, air quality data collection is performed synchronously, recording environmental parameters along the route to form a raw air quality dataset and ensuring the completeness of the collected information. For the raw air quality dataset, feature analysis methods are used to extract significant patterns in pollution distribution. A support vector machine algorithm is then used to classify the data and determine the distribution characteristics of potential pollution areas. If the pollution distribution characteristics indicate an abnormal area, real-time data and location positioning technology are combined to analyze the possible sources of temporary pollution and obtain the initial coordinate range of the pollution source. Using this coordinate range, the inspection robot is driven to approach the abnormal area to acquire more refined air quality information, update the detailed characteristics of the pollution distribution, and determine the specific location of the temporary pollution source. Based on the updated pollution distribution characteristics, the location positioning results are verified. If the verification results are inconsistent with the initial coordinate range, the movement drive of the inspection route is readjusted to acquire supplementary data and determine the final location of the pollution source.

[0052] The specific implementation method for the real-time updating of dynamic inspection paths, robot movement, air quality data collection, pollution source distribution feature extraction, and potential pollution source location is as follows: First, based on real-time environmental data and historical inspection records, the inspection path is updated using a dynamic programming algorithm. Assuming the initial path is from point A to point B, a distance of 10.5 kilometers, and the sensor detects a sudden increase in wind speed to 5.2 meters per second near point C, the system automatically adjusts the path to bypass point C, adding point D as a transit point, increasing the total path length to 12.3 kilometers. The optimized path is transmitted to the inspection robot control module via a wireless network, driving the robot to move along the path at a constant speed of 0.5 meters per second to ensure coverage of key areas. Secondly, during the movement, the air quality sensor onboard the robot collects data every 10 seconds, including PM2.5 and SO2 concentrations. Assuming a peak PM2.5 concentration of 85.6 micrograms per cubic meter is collected along a certain path, exceeding the normal value by 50 micrograms per cubic meter, the system automatically records the coordinates of this point as (X:23.4, Y:45.6) and marks it as an anomaly. Subsequently, the data analysis module performs spatial interpolation on the collected data, using Kriging interpolation to calculate the pollution concentration gradient. It is found that the concentration increases from coordinates (X:23.4, Y:45.6) towards the northwest, with a gradient value of 3.2 micrograms per cubic meter per meter. Combined with wind direction data (northwest wind, 3.1 m / s), it is inferred that the pollution source may be located 1.2 kilometers northwest of this point. Finally, based on gradient analysis and the wind direction model, the system constructs a pollution diffusion simulation, using a Gaussian plume model to estimate the location of the pollution source, concluding that the potential temporary pollution source is located at coordinates (X:22.2, Y:46.8).

[0053] The resource scheduling module is used to move and collect data based on the dynamic inspection path to determine the location of the pollution source;

[0054] Based on the location information of potential temporary pollution sources, the priority distribution of monitoring points is adjusted, scheduling instructions for multi-device collaborative operation are obtained, and if the pollution source characteristic value of a certain area exceeds the preset threshold, more inspection resources are preferentially allocated to that area.

[0055] By acquiring the location information and characteristic values ​​of pollution sources from the environmental monitoring system, a dynamic distribution map of regional divisions is constructed to obtain a preliminary pollution distribution status. Based on this distribution status, the characteristic values ​​of each regional division are compared with preset thresholds. If the characteristic value of a certain region exceeds the preset threshold, that region is marked as a high-priority region, determining the monitoring scope requiring focused attention. Using the distribution data of high-priority regions, combined with location information, monitoring priorities are adjusted to generate a targeted inspection task allocation plan and obtain preliminary instructions for multi-device collaborative operations. Based on these preliminary instructions, the resource allocation requirements for equipment collaborative operations are analyzed. Considering the urgency of the inspection tasks, the number of prioritized devices and route planning are optimized to obtain a refined scheduling plan. Based on this refined scheduling plan, real-time data on temporary pollution trends is acquired. Combined with the dynamic updates of regional divisions, the risk of pollution source spread is assessed. If the spread risk exceeds a preset threshold, the resource allocation plan is readjusted to determine new inspection priorities.

[0056] Specifically, the implementation method for adjusting the priority distribution of monitoring points based on the location information of potential temporary pollution sources and obtaining multi-device collaborative operation scheduling instructions can be achieved through the following logic and algorithm. First, assume that the system collects data in real time from 10 monitoring points within an industrial park via a sensor network. Each monitoring point has a coverage radius of 500 meters. The pollution source characteristic value is the PM2.5 concentration in the air (unit: micrograms per cubic meter), with a preset threshold of 75 micrograms per cubic meter. The system uses a GIS geographic information system to accurately locate potential pollution sources. Combined with historical data analysis, it identifies three high-risk areas: Area A has a PM2.5 concentration of 80 micrograms per cubic meter, exceeding the threshold; Area B has 60 micrograms per cubic meter; and Area C has 50 micrograms per cubic meter. Based on this, the system uses a priority allocation algorithm (e.g., weighted scoring method) to elevate the monitoring priority of Area A to the highest level, setting its weight to 0.5. The weights for Areas B and C are 0.3 and 0.2, respectively. This calculation shows that Area A needs to be allocated 50% of the inspection resources. Subsequently, the system invokes the multi-device collaborative scheduling module. Assuming there are 5 drone inspection devices and 10 ground monitoring robots in the park, the system calculates the shortest path and resource allocation scheme through optimization algorithms (such as genetic algorithms) to ensure that 3 drones and 6 robots prioritize coverage of area A, increasing the inspection frequency from once a day to three times a day. Simultaneously, scheduling instructions are generated: drones 1 to 3 perform gridded scanning in area A, with a grid unit of 50 meters × 50 meters; robots 1 to 6 are deployed along the boundary of area A, setting up a monitoring point every 100 meters. Data is uploaded to the cloud analysis platform in real time. If the pollution level in area A continues to exceed the threshold, the system dynamically adjusts the resource allocation ratio, further increasing the inspection frequency to five times a day. Through data analysis, the system predicts the pollution spread trend and, combined with wind speed and direction data (assuming a wind speed of 5 meters per second and a northeast direction), predicts that pollution may affect downstream area D, and proactively raises the monitoring priority of area D to a weight of 0.4.

[0057] The priority adjustment module is used to adjust the priority of monitoring points and allocate inspection resources based on the location of the pollution source, and generate inspection instructions.

[0058] By using scheduling instructions for multi-device collaborative operations, task allocation data for each inspection robot is transmitted. Target area and path planning information for each device are extracted from the task allocation data to obtain a coordinated optimization scheme for overall inspection coverage.

[0059] By employing a multi-device collaborative scheduling instruction set, the task allocation numbers for inspection robots are obtained from the data transmission chain, determining the initial task distribution for each device. Based on the initial task distribution, combined with target area points and area division methods, path planning maps for each inspection robot are generated, resulting in a preliminary inspection range allocation. Using the technical interface of the device interaction layer, real-time data interaction is performed on the preliminary inspection range allocation to determine if there are overlapping or blank areas. If overlapping areas are detected, the path planning map is adjusted; if blank areas are detected, the task allocation numbers for inspection robots are supplemented, determining the adjusted inspection range. Through the overall inspection network topology, the matching degree between the adjusted inspection range and coverage coordination is analyzed, obtaining information on the distribution of areas with insufficient coordination, resulting in a list of areas to be optimized. Based on the list of areas to be optimized, historical data from the optimization scheme library is invoked, combined with the task allocation numbers and path planning maps, to generate targeted optimization and adjustment strategies, determining the updated inspection coverage scheme. Through the scheduling instruction set, the updated inspection coverage scheme is distributed to each inspection robot, synchronously updating the data records of the device interaction layer to obtain the final collaborative operation status.

[0060] Specifically, in the process of implementing multi-device collaborative operation scheduling instructions to optimize inspection coverage, the central scheduling system first generates task allocation data. Assuming there are 5 inspection robots, the system allocates tasks based on the total inspection area of ​​5000 square meters and the priority of each area: Robot 1 is responsible for area A (1000 square meters), Robot 2 for area B (1200 square meters), Robot 3 for area C (800 square meters), Robot 4 for area D (900 square meters), and Robot 5 for area E (1100 square meters). Next, the system uses the A* algorithm to generate path planning information for each robot. For example, Robot 1's path planning in area A needs to pass through 10 key points, with a total path length of 200 meters. The algorithm calculates the Euclidean distance between each point (e.g., the distance from point 1 to point 2 is 15.5 meters) and combines it with obstacle avoidance weights (weight value 0.8) to optimize the shortest path. Subsequently, the system extracts target area and path planning information from the task allocation data. By analyzing the path length and area coverage of each robot (e.g., robot 1 has a coverage of 98%), the system calculates the overall inspection coverage coordination index and avoids path intersection conflicts (the conflict threshold is set at 5 meters), thereby forming an optimization plan.

[0061] The collaborative optimization module is used to optimize the overall inspection coverage coordination based on the inspection instructions and generate a coordination optimization scheme.

[0062] Based on the overall coordination optimization plan for inspection coverage, the dynamic paths and monitoring tasks of each inspection robot are updated in real time. The comprehensiveness index of regional coverage is extracted from the updated path data, and if a certain area is not covered, a supplementary inspection task is triggered.

[0063] The system acquires real-time location data and current path planning information of the inspection robot. This data is compared with preset area division information to determine if any uncovered areas exist. If the comparison reveals uncovered areas, coverage indicator data is extracted from the path information and compared using preset thresholds to determine the specific range and priority ranking of the uncovered areas. Based on the range and priority ranking of the uncovered areas, the inspection robot's path planning is dynamically adjusted, generating new path information and assigning supplementary tasks to higher-priority areas. By distributing the new path information to each inspection robot, the supplementary task execution process is initiated, acquiring real-time monitoring data during task execution. Area integrity-related indicators are extracted from the real-time monitoring data to determine if the supplementary task has covered the target area. If the preset coverage standard is not met, the path planning is readjusted.

[0064] Specifically, in optimizing the dynamic paths and monitoring tasks of the inspection robots, the location and task execution status of each robot are first obtained through a real-time data acquisition system. Assuming there are 5 inspection robots distributed in an industrial park of 1,000 square meters, each robot uploads its location coordinates and task completion data once per minute. The path planning module based on the A* algorithm is used to dynamically adjust the path and calculate the distance and estimated time from the current position to the next monitoring point for each robot. For example, robot 1 is currently located at coordinates (10,20), and the target point is (50,60). The shortest path calculated by the A* algorithm is 45 meters, and the estimated time is 3 minutes. At the same time, the task allocation is optimized by combining the task priority weight (e.g., the weight of emergency tasks is 0.8, and the weight of regular tasks is 0.2) to ensure that high-priority areas are covered first. Next, the system extracts the comprehensiveness index of area coverage from the updated path data. The system divides the park into 50 small areas of 10 square meters each, and counts the number of small areas covered by each robot's path. Assuming the current coverage rate is 85%, meaning 42 small areas have been inspected and 8 small areas remain uncovered, the system calculates and generates a coverage report in real time using the coverage rate formula (number of covered small areas / total number of small areas * 100%). Then, for the uncovered areas, the system automatically triggers supplementary inspection tasks. Based on the center point coordinates of the uncovered area (e.g., the center point of small area 43 is (70, 80)) and the robot's current load, it calculates the nearest idle robot (e.g., robot 2, located at (60, 70), 10 meters away, with a load rate of 30%) and assigns tasks, generating supplementary paths. This process is expected to take 1.5 minutes, ensuring coverage reaches 100%. To form a tight logical chain, the system also works in conjunction with the energy management module to analyze the expected increase of 5% in the power consumption of robot 2 after the supplementary task. If the increase is less than the safety threshold of 20%, the system will plan the charging path to the nearest charging station (coordinates (65,75)) in advance to ensure the continuity of the task.

[0065] The coverage detection module is used to update the inspection path and tasks based on the coordination optimization scheme and trigger supplementary inspection tasks.

[0066] For supplementary inspection tasks, the specific location information of the uncovered areas is obtained, and the coverage path is regenerated through path planning algorithm. Inspection time and resource consumption data are extracted from the generated path to determine the final inspection execution plan.

[0067] Location data of uncovered areas is acquired, and a Geographic Information System (GIS) is used to scan and analyze the target area to identify the extent of areas not covered by inspection and determine their specific distribution information. Based on this distribution information, a coverage path is generated using a path planning algorithm. Dijkstra's algorithm is used to calculate the shortest path from the current inspection point to the uncovered area, resulting in optimized coverage path data. Based on the coverage path data, inspection time information is extracted. Time estimation is performed by combining path length and inspection equipment speed to determine if the inspection time exceeds a preset threshold. If it does, the path planning is readjusted, resulting in adjusted time data. Using the adjusted time data, resource consumption is analyzed to obtain energy consumption and material usage data for the inspection equipment along the coverage path. If resource consumption exceeds a preset limit, the path is locally optimized to determine the final resource consumption data. Based on the final resource consumption data and time data, an inspection execution plan is generated. Task allocation logic is used to assign the coverage path to different inspection equipment, resulting in a specific task allocation scheme. Based on the task allocation plan and regional analysis data, the inspection execution plan is verified. If it is found that some areas are still not covered, the location data is reacquired and the path is adjusted to determine the final execution plan.

[0068] The task generation module is used to generate a coverage path and determine an inspection execution plan based on the supplementary inspection task.

[0069] Specifically, for supplementary inspection tasks, the target area is first analyzed using a Geographic Information System (GIS) in a grid pattern. Assuming a total area of ​​100 square kilometers, with 70 square kilometers already covered and 30 square kilometers uncovered, spatial analysis algorithms are used to identify the specific locations of the uncovered areas. For example, the uncovered areas are found to be concentrated within a 10-square-kilometer area in the northeast corner, with specific coordinates ranging from 120.5 to 120.8 degrees east longitude and 30.2 to 30.5 degrees north latitude. Next, based on the A* path planning algorithm, combined with terrain data and obstacle information (such as rivers and buildings), an optimal inspection path covering the uncovered areas is generated. Assuming a total path length of 50 kilometers, the starting point is the boundary point of the covered area (120.4 degrees east longitude, 30.1 degrees north latitude), and the ending point is the center point of the uncovered area (120.6 degrees east longitude, 30.3 degrees north latitude). The path planning considers the shortest distance and lowest energy consumption, and the calculated path includes 5 key nodes. Subsequently, inspection time and resource consumption data were extracted from the generated path. Assuming the average speed of the inspection equipment was 5 km / h, the total inspection time was 10 hours, and resource consumption, including electricity consumption, was 50 kWh, historical data analysis showed that the peak electricity consumption occurred at node 3 (120.55°E, 30.25°N), with a slope of 15 degrees, accounting for 30% of the total consumption. Finally, by comprehensively analyzing the time and resource data, the final inspection execution plan was determined. The system automatically compared three alternative plans (Plan A: 9 hours, 60 kWh consumption; Plan B: 10 hours, 50 kWh consumption; Plan C: 11 hours, 45 kWh consumption). Based on a weighted score of time and energy consumption (time accounts for 60%, energy consumption accounts for 40%), Plan B was calculated to have the highest score of 85 points (Plan A: 80 points; Plan C: 82 points). Therefore, Plan B was selected as the final execution plan.

[0070] The above are merely preferred embodiments of the present application, but the scope of protection of the present application is not limited thereto. Any changes or substitutions that can be easily conceived by a person skilled in the art within the technical scope disclosed in this application should be included in the scope of protection of the present application. Therefore, the scope of protection of the present application should be based on the scope of protection of the claims.

Claims

1. An indoor air quality inspection robot, characterized in that, include: The environmental perception module is used to acquire information on the indoor space layout and obstacle distribution; The area division module is used to divide the inspection area and determine the monitoring priority based on the indoor space layout and obstacle distribution information. The data acquisition module is used to collect air quality data in the inspection area and extract abnormal fluctuation characteristics to obtain the range of key inspection target areas; The dynamic monitoring module is used to acquire the current location and obstacle dynamic information based on the scope of the key inspection target area, and adjust the local path accordingly; The path planning module is used to generate a dynamic inspection path based on the local path; The resource scheduling module is used to move and collect data based on the dynamic inspection path to determine the location of the pollution source; The priority adjustment module is used to adjust the priority of monitoring points and allocate inspection resources based on the location of the pollution source, and generate inspection instructions. The collaborative optimization module is used to optimize the overall inspection coverage coordination based on the inspection instructions and generate a coordination optimization scheme. The coverage detection module is used to update the inspection path and tasks based on the coordination optimization scheme and trigger supplementary inspection tasks. The task generation module is used to generate a coverage path and determine an inspection execution plan based on the supplementary inspection task.

2. The robot according to claim 1, characterized in that, The environmental perception module extracts spatial layout and obstacle distribution data from the indoor environment through a pre-built 3D model, generates an initial inspection area division map, and uses data extraction technology to classify the spatial layout and obstacle distribution data, mark high-complexity areas, and determine the monitoring priority of high-complexity areas.

3. The robot according to claim 1, characterized in that, The data acquisition module collects data from each region through a sensor network to obtain the raw data set of air quality, and uses real-time analysis technology to continuously track the air quality change trend of each region, extract the change characteristics in the time series, and determine potential abnormal fluctuation ranges.

4. The robot according to claim 1, characterized in that, The dynamic monitoring module obtains the current location information and surrounding obstacle distribution data of the target area from the sensors carried by the inspection robot through a real-time data transmission mechanism, and uses a pre-established obstacle recognition model to analyze whether there are characteristics of temporary obstacles in the obstacle distribution data, and to determine whether there is a risk of obstruction.

5. The robot according to claim 1, characterized in that, The path planning module acquires environmental data and obstacle information from the current location, combines it with the coordinate data of the target area, constructs an initial path planning model, integrates real-time environmental change data of the local path, and uses adjustment strategies to dynamically correct the path to determine the feasibility of the local path.

6. The robot according to claim 1, characterized in that, The resource scheduling module drives the inspection robot to move along the designated path based on the real-time update results of the dynamic inspection path. At the same time, it collects air quality data along the way, extracts significant patterns of pollution distribution using feature analysis methods, and classifies the data using the support vector machine algorithm to determine the distribution characteristics of potential pollution areas.

7. The robot according to claim 1, characterized in that, The priority adjustment module obtains the location information and characteristic values ​​of pollution sources from the environmental monitoring system, constructs a dynamic distribution map of regional divisions, and compares the characteristic values ​​of each regional division with preset thresholds based on the pollution distribution status, marks high-priority areas, and determines the monitoring scope that needs to be focused on.

8. The robot according to claim 1, characterized in that, The collaborative optimization module obtains the task allocation number of the inspection robot from the data transmission chain through the multi-device collaborative scheduling instruction set, determines the initial task distribution of each device, and generates the path planning map of each inspection robot by combining the target area point and the area division method, thus obtaining the preliminary inspection range allocation.

9. The robot according to claim 1, characterized in that, The coverage detection module acquires the real-time location data and current path planning information of the inspection robot. By analyzing the data and comparing it with the preset area division information, it determines whether there are any uncovered areas and dynamically adjusts the path planning of the inspection robot according to the range and priority of the uncovered areas.

10. The robot according to claim 1, characterized in that, The task generation module acquires the location data of the uncovered area, uses a geographic information system to scan and analyze the target area, identifies the area not covered by the inspection, generates a coverage path through a path planning algorithm, and uses the Dijkstra algorithm to calculate the shortest path from the current inspection point to the uncovered area, thus obtaining the optimized coverage path data.

Images