Intelligent control method and system of air purification equipment

Abstract

The application provides an intelligent control method and system of an air purification device. In the application, multi-dimensional environmental data in an open pollution environment is obtained, including three-dimensional space data of pollutants captured by the air purification device, real-time air flow velocity and concentration distribution, a pollutant distribution heat map is generated, and the multi-dimensional data and the heat map are jointly coded to form a pollution concentration gradient atlas. At the same time, historical control strategy data sets are collected, cross-environment matching is performed with the atlas, and flow guiding control features associated with the diffusion direction of the pollutants are extracted. Based on the mapping relationship between the features and the atlas, a dynamic flow guiding parameter set is determined, and finally an intelligent control instruction of the air purification device is generated, directional air flow disturbance and air volume distribution optimization control of high concentration areas are realized. The technical scheme provided by the application can improve the efficiency and accuracy of intelligent control of the air purification device.

Description

Technical Field

[0001] This application relates to the field of intelligent equipment control technology, and in particular to an intelligent control method and system for an air purification device. Background Technology

[0002] In open, multi-source pollution environments, such as subway stations, large supermarkets, or farmers' markets—places with high pedestrian traffic and highly dynamic pollution source distribution—air pollution is characterized by its sudden onset, complex diffusion paths, and unpredictability. Pollutants may originate from various activities and infiltration from the external environment, and their concentrations vary drastically over time and space. In such environments, air purification equipment must be able to quickly identify the location of pollution sources, predict pollutant migration paths, and dynamically adjust purification strategies to cover changing pollution areas, thereby ensuring continuous and stable air quality.

[0003] A current mainstream solution is a real-time control system for air purification equipment based on big data analytics and dynamic optimization algorithms. This system deploys a high-density sensor network to collect real-time data on local environmental parameters such as pollutant concentration, airflow velocity, temperature, and humidity, and then combines this data with historical pollution data to construct a dynamic optimization model. The system utilizes cloud computing capabilities to perform correlation analysis on multi-source data, predict pollutant diffusion trends, and generate optimized equipment operating parameters. These parameters are then transmitted to terminal devices via wireless communication modules, enabling proactive intervention in polluted areas. This data-driven approach enhances the intelligence of control and is suitable for the initial treatment of certain dynamic pollution scenarios.

[0004] While the scheme enhances predictive capabilities through big data analytics, its core logic still relies on historical data modeling, making it ill-equipped to handle sudden pollution events or nonlinear diffusion phenomena under complex meteorological conditions. For example, in the event of sudden localized pollution in a subway station or strong winds, the system's predictive model, lacking the ability to correlate real-time weather with the sudden pollution source, is prone to delays or insufficient coverage in control strategies. Furthermore, the system's response to the synergistic effects of multiple pollution sources is weak, making it susceptible to misjudgments when pollution sources overlap or cross-spread, thereby affecting overall purification efficiency and the rationality of energy utilization. Summary of the Invention

[0005] This application provides an intelligent control method and system for air purification equipment to solve the problems of low efficiency and poor accuracy in the intelligent control of air purification equipment in the prior art.

[0006] In a first aspect, this application provides an intelligent control method for an air purification device, comprising:

[0007] Acquire multidimensional environmental data of pollutants in an open polluted environment, including three-dimensional spatial data of the pollutants, real-time airflow velocity, and pollutant concentration distribution data captured by air purification equipment;

[0008] A pollutant distribution heat map is generated based on the three-dimensional spatial data, and the multi-dimensional environmental data and the pollutant distribution heat map are jointly encoded to generate a pollution concentration gradient map.

[0009] A dataset of control strategies for historical polluted environments is collected synchronously, and the dataset is matched across environments with the pollution concentration gradient map to extract diversion control features that are spatially correlated with the pollutant diffusion direction in the pollution concentration gradient map.

[0010] Based on the correlation mapping relationship between the flow control characteristics and the pollution concentration gradient map, the dynamic flow control parameter set in the open pollution environment is determined;

[0011] The intelligent control instructions for the air purification device are generated based on the dynamic flow guidance parameter set. The intelligent control instructions include a directional airflow disturbance strategy and an airflow distribution scheme for the high-concentration area in the pollution concentration gradient map.

[0012] Optionally, the step of performing cross-environment matching between the control strategy dataset and the pollution concentration gradient map to extract diversion control features that are spatially correlated with the pollutant diffusion direction in the pollution concentration gradient map includes:

[0013] Based on the spatial vector distribution of pollutant diffusion direction in the pollution concentration gradient map, a subset of candidate control strategies that have a similarity to the spatial vector distribution reaching a preset threshold are selected from the control strategy dataset.

[0014] Spatial correlation analysis is performed on the deformation angle of the guide structure and the air volume distribution parameters in the candidate control strategy subset to establish a dynamic matching relationship between the guide control parameters and the pollutant diffusion direction in the control strategy dataset.

[0015] Based on the dynamic matching relationship, the combination of diversion control parameters whose spatial vector angle of the pollutant diffusion direction meets the preset conditions is extracted from the subset of candidate control strategies. Based on the spatial coverage of the diversion control parameter combination in the pollution concentration gradient map, the dynamic priority weight value of each diversion control parameter combination is calculated.

[0016] The flow control parameter combination is fused and reconstructed based on the dynamic priority weight value to generate flow control features that are spatially correlated with the direction of pollutant diffusion.

[0017] Optionally, the step of performing spatial correlation analysis on the deformation angle of the diversion structure and the airflow distribution parameters in the subset of candidate control strategies, and establishing a dynamic matching relationship between the diversion control parameters in the control strategy dataset and the pollutant diffusion direction, includes:

[0018] The deformation angle of the flow guiding structure in the candidate control strategy subset is spatially projected and compared with the direction of pollutant diffusion to generate a first geometric correlation feature;

[0019] The air volume allocation parameters in the candidate control strategy subset are matched with the spatial coverage of the corresponding pollution concentration gradient to generate a second geometric correlation feature.

[0020] Based on the superposition relationship between the first geometric association feature and the second geometric association feature, a dynamic matching rule base for the diversion control parameters and the pollutant diffusion direction in the control strategy dataset is established.

[0021] Based on the spatial vector distribution of pollutant diffusion direction in the pollution concentration gradient map, mapping relationships that meet preset conditions are selected from the dynamic matching rule base to obtain the selected mapping relationship set;

[0022] Based on the selected mapping relationship set, the deformation angle of the flow guiding structure and the air volume distribution parameters are bound together according to the spatial correlation strength to generate a dynamic matching relationship between the flow guiding control parameters and the pollutant diffusion direction.

[0023] Optionally, determining the dynamic flow guidance parameter set in the open pollution environment based on the correlation mapping relationship between the flow guidance control features and the pollution concentration gradient map includes:

[0024] Extract the combination of diversion control parameters associated with the pollutant diffusion direction in the pollution concentration gradient map from the diversion control features;

[0025] Based on the spatial vector distribution of pollutant diffusion direction in the pollution concentration gradient map, the reverse compensation intensity value between the deformation angle of the diversion structure and the spatial vector distribution in the diversion control parameter combination is calculated.

[0026] Based on the reverse compensation intensity value, spatial coverage priority coefficients are assigned to the diversion control parameter combination, and the diversion control parameter combination is sorted based on the spatial coverage priority coefficients to filter the target parameter subset that covers the high concentration area in the pollution concentration gradient map.

[0027] A dynamic flow guidance parameter set is generated based on the spatial synergistic effect strength between the flow guidance structure deformation angle and the air volume distribution parameter in the target parameter subset.

[0028] Optionally, the step of jointly encoding the multidimensional environmental data and the pollutant distribution heatmap to generate a pollution concentration gradient map includes:

[0029] Based on the pollution concentration gradient of the pollutants in the pollutant distribution heatmap, the migration direction of the pollutant particles is calculated, and based on the pollution concentration values ​​at each spatial location in the pollutant distribution heatmap, the dynamic correlation between the migration direction of the pollutant particles and the real-time airflow velocity in the multidimensional environmental data is determined.

[0030] Based on the dynamic correlation, the predicted diffusion path of the pollutant in three-dimensional space under the influence of the real-time airflow velocity is calculated, and based on the spatial distribution relationship between the predicted diffusion path and the pollutant concentration value, a real-time interactive mapping relationship between the pollutant concentration gradient and the real-time airflow velocity is established.

[0031] Through the real-time interactive mapping relationship, the static concentration distribution of the pollutant distribution heatmap is dynamically superimposed with the real-time airflow velocity to generate a pollution concentration gradient map with temporal continuity.

[0032] Optionally, calculating the predicted diffusion path of the pollutant in three-dimensional space affected by the real-time airflow velocity based on the dynamic correlation includes:

[0033] Based on the dynamic correlation, the spatial vector of the migration direction of the pollutant particles is superimposed with the vector of the real-time airflow velocity in three-dimensional space to generate the initial diffusion path prediction vector set of the pollutants.

[0034] Based on the pollution concentration values ​​at each spatial location in the pollutant distribution heatmap, the directions of each vector in the initial diffusion path prediction vector set are corrected by concentration gradient to obtain the diffusion path prediction results.

[0035] Optionally, generating intelligent control commands for the air purifier based on the dynamic airflow parameter set includes:

[0036] Based on the spatial distribution density of high-concentration areas in the pollution concentration gradient map and the real-time rate of change of the pollutant diffusion direction, dynamic execution weight values ​​are assigned to the dynamic diversion parameter set to generate the execution parameter set.

[0037] Based on the physical deformation range limitation of the adjustable airflow guiding structure of the air purification device, the deformation angle of the airflow guiding structure in the set of execution parameters is corrected by boundary constraints to obtain the corrected deformation angle of the airflow guiding structure.

[0038] The modified airflow guiding structure deformation angle and the corresponding airflow distribution parameters are matched and combined according to the spatial coverage density of the dynamic airflow guiding parameter set to generate intelligent control commands.

[0039] Secondly, this application provides an intelligent control system for an air purification device, comprising:

[0040] The acquisition module acquires multidimensional environmental data of pollutants in an open polluted environment. The multidimensional environmental data includes three-dimensional spatial data of the pollutants, real-time airflow velocity, and pollutant concentration distribution data captured by air purification equipment.

[0041] The encoding module generates a pollutant distribution heat map based on the three-dimensional spatial data, and jointly encodes the multi-dimensional environmental data with the pollutant distribution heat map to generate a pollution concentration gradient map.

[0042] The matching module synchronously collects a dataset of control strategies for historical polluted environments and performs cross-environment matching between the control strategy dataset and the pollution concentration gradient map to extract diversion control features that are spatially correlated with the pollutant diffusion direction in the pollution concentration gradient map.

[0043] The mapping module determines the dynamic flow guidance parameter set in the open pollution environment based on the correlation mapping relationship between the flow guidance control features and the pollution concentration gradient map.

[0044] The generation module generates intelligent control instructions for the air purification device based on the dynamic airflow parameter set. The intelligent control instructions include a directional airflow disturbance strategy and an airflow distribution scheme for the high-concentration area in the pollution concentration gradient map.

[0045] Thirdly, embodiments of this application provide a computing device, including a processing component and a storage component; the storage component stores one or more computer instructions; the one or more computer instructions are invoked and executed by the processing component to implement an intelligent control method for an air purification device as described in the first aspect above.

[0046] Fourthly, embodiments of this application provide a computer storage medium storing a computer program, which, when executed by a computer, implements an intelligent control method for an air purification device as described in the first aspect.

[0047] In this embodiment, multidimensional environmental data of pollutants in an open polluted environment is acquired. This multidimensional environmental data includes three-dimensional spatial data of the pollutants captured by an air purification device, real-time airflow velocity, and pollutant concentration distribution data. A pollutant distribution heatmap is generated based on the three-dimensional spatial data, and the multidimensional environmental data and the pollutant distribution heatmap are jointly encoded to generate a pollution concentration gradient map. A historical pollution environment control strategy dataset is simultaneously collected, and the control strategy dataset is cross-environmentally matched with the pollution concentration gradient map to extract flow guidance control features spatially correlated with the pollutant diffusion direction in the pollution concentration gradient map. Based on the correlation mapping relationship between the flow guidance control features and the pollution concentration gradient map, a dynamic flow guidance parameter set in the open polluted environment is determined. Intelligent control instructions for the air purification device are generated based on the dynamic flow guidance parameter set. These intelligent control instructions include directional airflow disturbance strategies and airflow allocation schemes for high-concentration areas in the pollution concentration gradient map.

[0048] The technical solution of this application has the following beneficial effects:

[0049] This application provides a high-precision dynamic data source for subsequent analysis by real-time acquisition of three-dimensional spatial distribution, airflow velocity, and concentration gradient data of pollutants, ensuring the comprehensiveness and real-time nature of pollution status perception. Multidimensional data is jointly encoded with heatmaps to construct a spatial-concentration correlation visualization model, intuitively revealing the pollution diffusion trend and concentration gradient change patterns. Through correlation analysis between historical data and current maps, the potential matching relationship between pollutant diffusion direction and historical diversion strategies is explored, improving the adaptability and transferability of control strategies. Based on the mapping relationship between features and maps, diversion parameters adapted to pollution diffusion characteristics are dynamically generated, achieving adaptive optimization of pollution control strategies. For high-concentration areas, directional airflow disturbance strategies and graded airflow distribution schemes are designed to precisely intervene in pollutant diffusion paths, improving the response efficiency of purification equipment to complex pollution scenarios.

[0050] Furthermore, during the cross-environment matching process, based on the pollutant diffusion vector direction in the pollution concentration gradient map, a subset of candidate strategies meeting the similarity criteria is selected from historical control strategies. Then, combinations of diversion parameters with suitable diffusion direction vector angles are extracted, and dynamic priority weights are calculated based on their spatial coverage in the map. Finally, diversion control features strongly correlated with the diffusion direction are generated through weight fusion and reconstruction. By screening for similarity between the pollution diffusion vector and historical strategies, and calculating the priority weights of parameter combinations, dynamic optimization and spatial adaptation of diversion strategies are achieved. This enhances the matching accuracy between control commands and pollutant diffusion paths, avoids redundant interference from historical data, and ensures the scientific nature of diversion feature extraction and the effectiveness of strategy deployment.

[0051] These or other aspects of this application will become more apparent in the following description of the embodiments. Attached Figure Description

[0052] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0053] Figure 1 A flowchart of an intelligent control method for an air purification device provided in this application is shown;

[0054] Figure 2 This application provides a schematic diagram of the structure of an intelligent control system for an air purification device.

[0055] Figure 3 A schematic diagram of the structure of a computing device provided in this application is shown. Detailed Implementation

[0056] To enable those skilled in the art to better understand the present application, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings.

[0057] In some of the processes described in the specification, claims, and accompanying drawings of this application, multiple operations appearing in a specific order are included. However, it should be clearly understood that these operations may not be executed in the order they appear herein, or may be executed in parallel. The operation numbers, such as 101, 102, etc., are merely used to distinguish different operations and do not themselves represent any execution order. Furthermore, these processes may include more or fewer operations, and these operations may be executed sequentially or in parallel. It should be noted that the descriptions such as "first," "second," etc., in this document are used to distinguish different messages, devices, modules, etc., and do not represent a chronological order, nor do they limit "first" and "second" to different types.

[0058] In open, multi-source pollution scenarios, such as subway stations, large supermarkets, or farmers' markets, air pollution is characterized by its sudden onset, complex diffusion paths, and unpredictability. While existing mainstream solutions achieve dynamic control through high-density sensor networks and historical data modeling, significant shortcomings remain. Modeling logic relying on historical data struggles to adapt to nonlinear diffusion phenomena under sudden pollution events or complex meteorological conditions, leading to lagging control strategies. Furthermore, the systems lack the ability to respond to the synergistic effects of multiple pollution sources, easily misjudging situations when pollution sources overlap or cross-diffusion occurs, reducing purification efficiency and energy utilization. In addition, existing systems have weak analysis of the dynamic correlation between real-time airflow disturbances and pollutant diffusion directions, making it difficult to accurately cover high-concentration pollution areas, resulting in suboptimal localized treatment effects.

[0059] To address the aforementioned shortcomings, this application proposes an intelligent control method for air purification equipment driven by multi-dimensional environmental data. By acquiring real-time data on the three-dimensional spatial distribution of pollutants, airflow velocity, and concentration, a pollution concentration gradient map is generated. Combined with spatial correlation analysis of historical control strategies and current diffusion paths, flow guidance control features are dynamically extracted, and a directional airflow disturbance strategy is generated. This solution overcomes existing technical bottlenecks through the following innovations: Utilizing multi-dimensional data joint encoding and cross-environment matching technology, it achieves real-time dynamic adaptation between pollutant diffusion direction and control strategy, significantly improving the response speed to sudden pollution events. Based on spatial vector correlation analysis, airflow allocation is optimized to accurately cover areas where multiple pollution sources intersect, avoiding purification blind spots caused by model lag or misjudgment. Finally, through the generation of dynamic flow guidance parameter sets and directional intervention in high-concentration areas, it effectively solves the core problems of insufficient pollution diffusion path prediction, weak multi-source collaborative control, and low local treatment efficiency in complex scenarios, providing an efficient and precise air purification solution for open environments.

[0060] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0061] Figure 1 A flowchart of an intelligent control method for an air purification device is provided in this application embodiment, such as... Figure 1 As shown, the method includes:

[0062] 101. Obtain multidimensional environmental data of pollutants in an open polluted environment, wherein the multidimensional environmental data includes three-dimensional spatial data of the pollutants, real-time airflow velocity, and pollutant concentration distribution data captured by air purification equipment;

[0063] In this step, multidimensional environmental data refers to the three-dimensional spatial coordinates of pollutants, real-time airflow velocity, and pollutant concentration distribution data.

[0064] Air purification equipment captures data in real time through multimodal sensors integrated into the purification equipment and uploads it to the central processing unit via an Internet of Things (IoT) communication module.

[0065] Air purification equipment is a device that removes pollutants from the air through physical filtration, chemical adsorption, ionization, or catalysis to improve air cleanliness.

[0066] An open-air polluted environment refers to a scenario where pollution sources are dynamically distributed and people move frequently, making the diffusion paths of pollutants complex and difficult to predict.

[0067] Three-dimensional spatial data refers to data that describes the position, shape, and distribution characteristics of an object or phenomenon in three-dimensional space using spatial coordinates.

[0068] Real-time airflow velocity refers to the quantified value of the direction and speed of airflow at a certain moment, usually expressed in vector form.

[0069] Pollutant concentration distribution data refers to a quantitative record of the changes in pollutant concentration at different locations within a certain area over time and space.

[0070] In this embodiment, a multimodal sensor network deployed on air purification equipment is used to collect real-time three-dimensional spatial data of pollutants, real-time airflow velocity, and pollutant concentration distribution data. First, a lidar sensor is used to perform a three-dimensional scan of the environment, generating point cloud data containing spatial coordinates and reflection intensity. This data is then combined with real-time localization and mapping algorithms to construct three-dimensional spatial data. Second, an ultrasonic anemometer is used to measure airflow velocity, and a filtering algorithm is used to eliminate noise interference, generating continuous time-series real-time airflow velocity data. Simultaneously, a distributed particulate matter sensor network is used to periodically collect pollutant concentration information, and a spatial interpolation algorithm is used to generate pollutant concentration distribution data. Finally, the three-dimensional spatial data, airflow velocity data, and pollutant concentration distribution data are encapsulated with a unified spatiotemporal label and transmitted to computing nodes via an IoT communication protocol, forming multi-dimensional environmental data and providing a dynamic data foundation for subsequent analysis.

[0071] A lidar sensor array is deployed inside the subway station to capture the three-dimensional spatial distribution of pollutants in real time, as well as the concentration of respirable particulate matter (PPM) in densely populated passenger areas. Airflow velocity sensors are installed at ventilation openings and platform edges to measure airflow direction and velocity. PPM sensors cover all channels, collecting concentration data. Multi-source data is integrated using timestamp alignment technology to generate a multi-dimensional environmental dataset containing spatial coordinates, airflow velocity vectors, and concentration values. During the morning rush hour, a sudden increase in PPM concentration near an exit is recorded; the system records its three-dimensional location and airflow velocity, creating a dynamic snapshot of the polluted environment, providing a basis for subsequent analysis.

[0072] 102. Generate a pollutant distribution heat map based on the three-dimensional spatial data, and jointly encode the multi-dimensional environmental data and the pollutant distribution heat map to generate a pollutant concentration gradient map;

[0073] In this step, the pollutant distribution heatmap refers to a two-dimensional / three-dimensional visualization model that uses color gradients to characterize the concentration of pollutants, with red indicating high concentration and blue indicating low concentration.

[0074] Joint coding refers to mapping multidimensional data into a unified mathematical expression and extracting key features through coding techniques.

[0075] Pollution concentration gradient map refers to a dynamic map that integrates spatial, concentration, and airflow parameters. Vector arrows indicate the direction of pollutant diffusion, and color depth indicates the rate of change of concentration gradient.

[0076] In this embodiment, firstly, a pollutant distribution heatmap is generated based on three-dimensional spatial data of pollutants using a spatial interpolation algorithm, visually displaying the spatial distribution characteristics of pollutant concentrations. Then, the pollutant distribution heatmap, real-time airflow velocity data, and pollutant concentration data are jointly encoded using a convolutional neural network to extract a multidimensional feature tensor. Finally, gradient calculation is performed on the encoded multidimensional feature tensor to quantify the gradient direction and intensity of pollutant diffusion, generating a pollution concentration gradient map.

[0077] The entrances and exits of the subway station exhibited red high-concentration zones. A pollutant distribution heatmap was generated using spatial interpolation algorithms based on 3D spatial data. This heatmap, along with airflow velocity data, was jointly encoded using a convolutional neural network to extract the gradient direction and intensity of pollutant diffusion. Pollutants diffused from the entrances and exits towards the center of the platform, with the gradient direction from northeast to southwest, and the concentration decreasing per meter. The final result was a pollution concentration gradient map, visually displaying the diffusion path and providing a basis for diversion strategies.

[0078] 103. Synchronously collect historical pollution control strategy datasets, and perform cross-environment matching between the control strategy datasets and the pollution concentration gradient map to extract diversion control features that are spatially correlated with the pollutant diffusion direction in the pollution concentration gradient map;

[0079] In this step, the direction of pollutant diffusion refers to the main paths and directions of pollutant migration and spread in the environment.

[0080] The historical control strategy dataset contains a database of the deformation angle of the airflow structure, airflow distribution parameters, and corresponding purification effects in historical scenarios.

[0081] Cross-environment matching refers to screening historical strategies that are similar to the current pollution diffusion pattern through similarity metrics and analyzing their parameter correlations.

[0082] The flow control characteristics are extracted from historical data and are a combination of baffle angle and airflow distribution ratio that are strongly correlated with the current diffusion direction.

[0083] In this embodiment, firstly, a dataset of control strategies for historical polluted environments, such as wind speed, wind direction, and airflow allocation schemes, is synchronously loaded from a cloud database. The spatial vector of the pollutant diffusion direction in the current gradient map is calculated, and its directional similarity is measured with the diffusion vectors in the historical dataset to select candidate control strategies that meet the similarity criteria. Secondly, a correlation analysis is performed on the deformation angle of the guiding structure and the airflow allocation parameters in the candidate strategies to construct a regression model between the guiding parameters and the purification effect, identifying key parameter combinations that significantly affect purification efficiency. Subsequently, based on the spatial coverage of the parameter combinations in the pollution concentration gradient map and historical purification efficiency, a weighted calculation method is used to evaluate the dynamic priority of each combination. Finally, the parameter combinations are fused and reconstructed based on the priority weights to generate guiding control features strongly correlated with the current pollutant diffusion direction, providing a basis for dynamic guiding parameter optimization.

[0084] This study investigates wind speed strategies and airflow distribution ratios during a morning rush hour at a subway station. A historical control strategy dataset is loaded from a cloud database. The matching degree between the current gradient map and historical strategies is calculated using a cosine similarity algorithm to select suitable strategies. Furthermore, principal component analysis is used to extract diversion control features, including the angle between the wind speed vector and the diffusion direction, and the airflow distribution ratio. Candidate diversion parameter combinations are generated to provide input for the dynamic diversion parameter set.

[0085] 104. Based on the correlation mapping relationship between the flow control characteristics and the pollution concentration gradient map, determine the dynamic flow control parameter set in the open pollution environment;

[0086] In this step, the association mapping relationship is the mathematical relationship between the flow control features and the gradient map parameters.

[0087] The dynamic airflow parameter set is a set of real-time control parameters generated based on the mapping relationship, including the airflow vane angle, airflow distribution ratio, and disturbance frequency.

[0088] In this embodiment, firstly, a mapping model is established using the random forest algorithm to connect airflow control features, such as wind speed vectors, airflow distribution ratios, and pollution concentration gradient maps. Then, based on the gradient direction and intensity of the current pollution concentration gradient map, dynamic airflow parameters, such as wind speed, wind direction, and airflow distribution ratios, are calculated using the mapping model. Next, a reinforcement learning algorithm is used to iteratively optimize the dynamic airflow parameters to ensure they effectively intercept pollutant diffusion paths. Finally, a set of dynamic airflow parameters with real-time adjustment capabilities is generated, providing precise input for equipment control.

[0089] In the subway station, based on the flow control features extracted in step 103, a mapping model between the flow control features and the pollution concentration gradient map is established using the random forest algorithm. Dynamic flow control parameters are calculated based on the gradient direction (northeast to southwest) and intensity of the current gradient map, with the main air duct wind speed in the northeast direction and the auxiliary air duct wind speed in the northwest direction. The parameters are iteratively optimized using a reinforcement learning algorithm, increasing the main air duct wind speed to enhance the interception effect. Finally, a dynamic flow control parameter set is generated for equipment control.

[0090] 105. Generate intelligent control instructions for the air purification device based on the dynamic flow guidance parameter set. The intelligent control instructions include a directional airflow disturbance strategy and an airflow distribution scheme for the high-concentration area in the pollution concentration gradient map.

[0091] In this step, intelligent control commands are control commands dynamically generated and executed by an intelligent system through artificial intelligence, the Internet of Things, and automation technologies, used to accurately and efficiently manage the operating status of equipment or systems.

[0092] Directional airflow disturbance strategy refers to disrupting the diffusion path of pollutants by periodically adjusting the angle of the deflector, thereby creating local turbulence and promoting mixing.

[0093] The air volume distribution scheme refers to the differentiated allocation of air volume at each air outlet based on the spatial distribution of high-concentration areas, forming a directional airflow barrier.

[0094] In this embodiment, firstly, the dynamic airflow parameter set is converted into intelligent control commands executable by the device. Based on the angle and frequency parameters in the airflow parameter set, a periodic oscillation mode of the airflow guide plate is designed. A proportional-integral-derivative controller is used to precisely control the oscillation angle, forming a directional airflow disturbance strategy. Secondly, based on the spatial distribution ratio of high-concentration areas, a fuzzy logic algorithm is used to dynamically calculate the airflow distribution scheme for each air outlet, adjusting the valve opening and fan speed. Finally, the control commands are encapsulated into standard communication protocol data packets and sent to the air purification device's execution end, driving the airflow guide mechanism and fan system to work together to form local turbulence that disrupts the pollutant diffusion path. Simultaneously, a directional airflow barrier is constructed to block the diffusion of high-concentration areas, ultimately achieving efficient dynamic purification of pollutants in open environments.

[0095] Inside the subway station, the system converts the dynamic airflow parameters—main duct wind speed in the northeast direction and auxiliary duct wind speed in the northwest direction—into executable control commands for the equipment. These commands control the fan speed and adjust the airflow direction using servo motors. For high-concentration areas near entrances and exits in the pollution concentration gradient map, a directional airflow disturbance strategy is generated to initiate local vortex airflow and reverse airflow velocity. The controller proportionally allocates airflow between the main and auxiliary ducts. After execution, the concentration of respirable particulate matter in high-concentration areas decreases. The system then fine-tunes the wind speed using algorithms to further optimize the purification effect, achieving proactive intervention and dynamic control of air pollution in the subway station.

[0096] In summary, steps 101 to 105, through real-time acquisition of multi-dimensional environmental data and gradient map modeling, combined with cross-environmental matching of historical strategies and generation of dynamic flow guidance parameters, achieve accurate prediction and proactive intervention of pollutant diffusion paths in open polluted environments. In complex scenarios such as subway stations, the system can quickly identify high-concentration areas and generate directional airflow disturbance strategies, improving the efficiency of pollutant concentration reduction while reducing energy consumption through optimized airflow allocation. Compared to traditional passive purification methods, this method significantly improves the real-time performance, adaptability, and energy efficiency of air purification, solving the challenges of managing cross-diffusion from multiple pollution sources and sudden pollution events.

[0097] To address the compatibility issue between historical control strategies and dynamic pollution scenarios, the solution employs cross-environment matching of control strategy datasets and pollution concentration gradient maps to extract diversion control features spatially correlated with the pollutant diffusion direction, thereby improving the matching accuracy between strategies and dynamic pollution characteristics. In some embodiments, step 103, which involves cross-environment matching of the control strategy dataset and the pollution concentration gradient map to extract diversion control features spatially correlated with the pollutant diffusion direction in the pollution concentration gradient map, includes:

[0098] 201. Based on the spatial vector distribution of pollutant diffusion direction in the pollution concentration gradient map, a subset of candidate control strategies that have a similarity to the spatial vector distribution reaching a preset threshold are selected from the control strategy dataset;

[0099] In step 201, the spatial vector distribution refers to the vector set composed of the pollutant diffusion direction and diffusion rate in the pollution concentration gradient map, used to describe the spatial migration trend of pollutants. The candidate control strategy subset refers to the set of historical diversion parameter combinations selected from the historical control strategy dataset that meet preset conditions in terms of similarity to the current pollutant diffusion direction. The preset threshold refers to a pre-set numerical value or conditional limit in a specific system, algorithm, or application scenario, used to determine whether a certain indicator has reached or exceeded a critical state, thereby triggering corresponding actions, alarms, or strategy adjustments.

[0100] In this embodiment, the spatial vector distribution of the pollution diffusion direction is first extracted, for example, by obtaining the airflow direction vector through simulation or sensor networks. Then, historical or preset diversion control parameters are extracted from the control strategy dataset and transformed into feature vectors corresponding to the spatial vector distribution. Using a cosine similarity algorithm, the feature vector of each diversion control parameter is matched with the spatial vector of the pollution diffusion direction to filter out a subset of candidate control strategies whose similarity reaches a preset threshold.

[0101] 202. Perform spatial correlation analysis on the deformation angle of the guide structure and the air volume distribution parameters in the candidate control strategy subset, and establish a dynamic matching relationship between the guide control parameters in the control strategy dataset and the pollutant diffusion direction;

[0102] In step 202, the airflow guiding structure deformation angle refers to the physical deflection angle of the airflow guide plate of the air purification equipment, used to adjust the airflow direction. Dynamic matching relationship refers to the mathematical correlation model between historical airflow guiding parameters and the current diffusion direction. Airflow distribution parameters refer to quantitative indicators used in ventilation or air conditioning systems to describe and regulate the airflow ratio, flow rate, and distribution strategy between different areas or equipment. Airflow control parameters refer to physical or algorithmic setpoints used to adjust the airflow direction, speed, or intensity to optimize pollutant diffusion or ventilation efficiency. Spatial correlation analysis refers to methods for studying the interdependencies between variables in geospatial data, revealing spatial patterns by quantifying the attribute similarities of neighboring areas or locations.

[0103] In this embodiment, firstly, spatial autocorrelation analysis is used to evaluate the spatial correlation between the deformation angle of the diversion structure and the airflow distribution parameters in the candidate strategy subset and the pollution diffusion direction. By constructing a mapping function between the diversion control parameters and the pollution vector, the dynamic matching relationship between the two is quantified. Furthermore, by combining real-time pollutant diffusion direction data, the diversion parameters are dynamically adjusted to form a dynamic matching model that adapts to environmental changes.

[0104] 203. Based on the dynamic matching relationship, extract the combination of diversion control parameters whose spatial vector angle of the pollutant diffusion direction meets the preset conditions from the subset of candidate control strategies, and calculate the dynamic priority weight value of each combination of diversion control parameters based on the spatial coverage of the diversion control parameter combination in the pollutant concentration gradient map.

[0105] In step 203, the spatial vector angle refers to the angular difference between the current diffusion direction vector and the corresponding vector of the flow guidance parameter in the historical strategy, used to evaluate the directional matching degree. The dynamic priority weight value refers to the comprehensive priority score calculated based on the spatial coverage and purification efficiency of the parameter combination. The flow guidance control parameter combination refers to the set of multiple parameters used to coordinately control the direction, speed, and intensity of airflow in a ventilation or airflow regulation system. Preset conditions refer to the initial premises or rules that are assumed to be in effect during system operation or decision-making. These conditions usually do not need to be explicitly stated but form the basis for subsequent operations. Spatial coverage refers to the area within space that a specific system or device can effectively operate or influence.

[0106] In this embodiment, firstly, based on dynamic matching relationships, combinations of flow control parameters with spatial vector angles in the pollutant diffusion direction smaller than a preset angle are selected. Secondly, the pollutant concentration gradient map is divided into gridded regions, and the effective coverage of each flow control parameter combination is statistically analyzed, for example, the percentage of grids covering high-concentration areas. Subsequently, combining purification efficiency indicators from historical data, such as the concentration decrease rate per unit time, the entropy weight method is used to calculate the weight value of each parameter combination. Finally, dynamic priority weight values ​​are generated according to the weight values ​​from high to low, ensuring that parameter combinations with broad coverage and high efficiency participate in subsequent fusion first.

[0107] 204. The flow control parameter combination is fused and reconstructed according to the dynamic priority weight value to generate flow control features that are spatially related to the direction of pollutant diffusion.

[0108] In step 204, fusion and reconstruction refers to weighting and superimposing parameter combinations according to priority to generate an optimal set of guiding parameters that adapts to the current diffusion direction. Spatial correlation refers to the interdependence or correlation between objects or phenomena in geographic space, which is usually manifested as the similarity or difference of attribute values ​​in neighboring areas or locations.

[0109] In this embodiment, firstly, based on dynamic priority weight values, the angles of the guide vanes in the candidate guide parameter combinations are weighted and averaged. For example, angles with higher weights account for a larger proportion, forming preliminary angle parameters. Secondly, a fuzzy clustering algorithm is used to merge airflow allocation patterns with similar strategies to eliminate redundant parameters. Subsequently, the effectiveness of the fused parameters is verified based on spatial coverage, and the weight allocation is adjusted to optimize coverage blind spots. Finally, a complete set of parameters, including guide vane angles, airflow allocation ratios, and disturbance frequencies, is generated that is strongly correlated with the current pollutant diffusion direction.

[0110] Here is a specific example:

[0111] In subway station pollution control, the system first identifies the current pollutant diffusion direction as northeast based on real-time monitoring data. Then, it selects candidate strategies from a historical strategy database that correspond to the angle between the current diffusion direction and the target direction, ensuring accurate direction matching. Further analysis reveals that the combination of the deflector angle and the main airflow volume shows the highest correlation with the diffusion direction offset, indicating that this parameter combination has a significant effect on suppressing pollution diffusion. Through spatial coverage assessment, this strategy covers the current high-concentration area. Combined with historical purification efficiency, dynamic priority calculation is performed to obtain a comprehensive priority weight. Based on this, the system integrates high-weight strategies to generate optimized deflection control characteristics. Adjusting the deflector angle and setting the main airflow volume are simultaneously sent to the purification equipment for execution, achieving dynamic adaptation to the pollution diffusion path and efficient treatment.

[0112] In summary, steps 201 to 204 utilize the spatial vector distribution of pollutant diffusion directions in the pollution concentration gradient map and employ a spatial vector similarity algorithm to select a subset of candidate control strategies that are highly consistent with the pollution diffusion direction, ensuring a precise spatial match between control parameters and diffusion directions. Furthermore, after deploying the strategies in complex scenarios, simulation verification and feedback from actual monitoring data demonstrate a significant reduction in pollutant concentrations and improved air quality, fully validating the practicality and scalability of this method in variable environments.

[0113] To address the insufficient modeling of the spatial correlation between diversion control parameters and pollutant diffusion direction, the scheme constructs a dynamic matching rule base by overlaying two sets of features, and filters mapping relationships based on the distribution of pollutant diffusion vectors. Finally, it binds the diversion angle and airflow parameters according to the strength of their spatial correlation, forming a dynamic matching relationship and achieving multi-dimensional collaborative optimization of control parameters and diffusion direction. In some embodiments, step 202, which involves performing spatial correlation analysis on the diversion structure deformation angle and airflow allocation parameters in the candidate control strategy subset to establish a dynamic matching relationship between the diversion control parameters and the pollutant diffusion direction in the control strategy dataset, includes:

[0114] 301. Spatial projection comparison of the deformation angle of the flow guiding structure in the subset of candidate control strategies with the direction of pollutant diffusion to generate a first geometric correlation feature;

[0115] In step 301, spatial projection comparison refers to comparing the airflow guidance direction corresponding to the deformation angle of the guide structure with the pollutant diffusion direction through geometric projection, quantifying the consistency of their directions. The deformation angle of the guide structure refers to the adjustable angle of the guide structure in terms of physical form or geometry, used to optimize the flow path or direction of the fluid. The first geometric correlation feature refers to the feature vector composed of the projection angle and the overlapping area, reflecting the ability of the guide angle to correct the diffusion direction.

[0116] In this embodiment, firstly, the deformation angles of the guiding structures in the candidate control strategy subset are converted into spatial vectors, and the pollutant diffusion direction is also converted into spatial vectors to ensure that both are in the same coordinate system. The spatial matching degree between the guiding structure and the diffusion direction is quantified by calculating the angular deviation and the ratio of the projected lengths of the two spatial vectors. Subsequently, the angular deviation and the ratio of the projected lengths are combined to generate a first geometric correlation feature, which is used to characterize the adaptability of the guiding parameters to the diffusion direction.

[0117] 302. Perform regional density matching between the air volume allocation parameters in the candidate control strategy subset and the spatial coverage of the corresponding pollution concentration gradient to generate a second geometric correlation feature;

[0118] In step 302, regional density matching refers to the consistency between the density of the airflow intensity distribution corresponding to the airflow allocation parameters and the density of the spatial coverage area of ​​the pollution concentration gradient. The second geometric correlation feature refers to the feature vector composed of the correlation coefficient between the airflow allocation ratio and the coverage density of the high-concentration area, reflecting the effect of airflow on the regulation of pollution concentration.

[0119] In this embodiment, firstly, the airflow allocation parameters are converted into coverage area simulation results, and high-concentration areas are extracted by combining the spatial coverage area of ​​the pollution concentration gradient map. Then, a regional density matching algorithm is used to analyze the overlap density between the airflow allocation coverage area and the high-concentration area. Subsequently, the coverage density is combined with historical purification efficiency weights to generate a second geometric correlation feature. This feature characterizes the coverage capability of the airflow parameters over the polluted area, providing a basis for coverage performance evaluation for dynamic matching.

[0120] 303. Based on the superposition relationship between the first geometric association feature and the second geometric association feature, establish a dynamic matching rule base between the diversion control parameters in the control strategy dataset and the pollutant diffusion direction;

[0121] In step 303, the superposition relationship refers to the weighted fusion of the first geometric correlation feature and the second geometric correlation feature to form a comprehensive matching rule. The dynamic matching rule base refers to a database containing correlation rules for guide angle, air volume distribution parameters, and pollutant diffusion direction.

[0122] In this embodiment, firstly, the first geometric association feature and the second geometric association feature are weighted and summed according to preset weights to generate a comprehensive matching degree. A threshold is set based on the comprehensive matching degree, and the diversion control parameters that meet the conditions are stored in the rule base, with applicable scenarios labeled. The rule base dynamically optimizes the matching strategy by updating the threshold or weights in real time, forming a dynamic matching rule base adapted to different diffusion directions.

[0123] 304. Based on the spatial vector distribution of pollutant diffusion direction in the pollution concentration gradient map, select mapping relationships that meet preset conditions from the dynamic matching rule base to obtain the set of selected mapping relationships;

[0124] In step 304, the mapping relationship set refers to the set of guiding parameter combinations selected from the dynamic matching rule base that are compatible with the spatial vector distribution of the current pollutant diffusion direction. The spatial vector distribution refers to a set of vectors arranged or distributed in space according to a certain pattern.

[0125] In this embodiment, firstly, the main vector of the current pollutant diffusion direction and its neighboring directions are extracted from the pollution concentration gradient map. Secondly, all rule entries matching the pollutant diffusion direction are retrieved from the dynamic matching rule base. Rule entries are filtered according to preset conditions, retaining highly adaptable strategies. Finally, a filtered mapping relationship set is generated, providing candidate strategies for subsequent parameter binding.

[0126] 305. Based on the selected mapping relationship set, the deformation angle of the flow guiding structure and the air volume distribution parameters are bound according to the spatial correlation strength to generate a dynamic matching relationship between the flow guiding control parameters and the pollutant diffusion direction.

[0127] In step 305, spatial correlation strength refers to the weighted binding of the guide angle and airflow distribution parameters according to the score of the parameter combination in the mapping relationship set, forming the final matching relationship. Dynamic matching relationship refers to the dynamic adjustment of the adaptability relationship between input parameters and output target in complex systems based on real-time environmental changes or target requirements, through algorithms or rule bases.

[0128] In this embodiment, firstly, the spatial correlation strength is calculated for each rule in the selected mapping relationship set. For example, rules with high overall matching degree are sorted in descending order of strength, and multiple high-weight strategies are selected, such as multiple combinations of guide structure deformation angles and airflow distribution parameters. A weighted average algorithm is then used to prioritize the strategy with the highest matching degree and fuse the binding parameters to generate the final combination of guide structure deformation angles and airflow distribution parameters. This parameter combination achieves the spatial correlation strength binding between the guide structure deformation angles and airflow distribution parameters, forming a dynamic matching relationship that dynamically adapts to the pollutant diffusion direction.

[0129] Here is a specific example:

[0130] In the pollution control of the subway station, real-time monitoring revealed that the pollutant diffusion direction was northeast. The system selected multiple candidate strategies from the historical strategy library, among which the combination of diversion angle and main air volume showed the highest correlation with the diffusion direction offset. The first geometric correlation feature of this combination was determined through spatial projection comparison. The second geometric correlation feature, determined through regional density matching, was coverage density and weight. In the dynamic matching rule library, the overall matching degree of this combination met the preset conditions. Finally, the system integrated the top high-weight strategies to generate diversion control parameters, which were then sent to the purification equipment for execution, achieving dynamic adaptation and efficient control of the pollution diffusion path.

[0131] In summary, steps 301 to 305 quantify the spatial correlation between flow guidance parameters and pollutant diffusion direction through geometric projection and regional density matching. Based on a dynamic rule base and parameter fusion technology, a high-precision matching relationship is generated, resolving the control deviation problem caused by direction-concentration decoupling in traditional methods. In subway station scenarios, the flow guidance angle matching error is reduced, the coverage of high-concentration areas is improved, and purification energy consumption is decreased.

[0132] To address the challenge of accurately adapting dynamic diversion parameters in open pollution environments, this solution generates a dynamic diversion parameter set through a reverse compensation mechanism. First, parameter combinations associated with the pollution diffusion direction are extracted from diversion control features, and spatial priority coefficients are assigned based on coverage area. A subset of target parameters covering high-concentration areas is then selected through priority ranking. Next, a dynamic diversion parameter set is generated based on the spatial synergistic effect strength of the parameter combinations, ensuring efficient interception and dilution of the pollution source. In some embodiments, step 104, determining the dynamic diversion parameter set in the open pollution environment based on the correlation mapping relationship between the diversion control features and the pollution concentration gradient map, includes:

[0133] 401. Extract the combination of diversion control parameters that are associated with the pollutant diffusion direction in the pollution concentration gradient map from the diversion control features;

[0134] In step 401, the flow control parameter combination refers to the set of parameters extracted from the flow control characteristics, such as the guide vane angle and airflow distribution ratio, which are spatially correlated with the current pollution diffusion direction. The pollutant diffusion direction refers to the main spatial path or trend of pollutant migration and propagation in the environment, which is usually determined by the combined effects of physical, chemical, and biological processes.

[0135] In this embodiment, firstly, all possible combinations of flow guidance structure deformation angles and airflow distribution parameters are extracted from the flow guidance control features, and these are correlated with the pollutant diffusion direction in the pollution concentration gradient map. A clustering algorithm is used to identify candidate combinations of parameters that have a high degree of matching with the diffusion direction. Subsequently, a Gaussian diffusion model is used to verify the pollutant interception efficiency of the candidate combinations, and flow guidance control parameter combinations with a strong correlation to the diffusion direction are selected. This combination directly reflects the adaptability of the flow guidance parameters to the diffusion direction, providing basic data for subsequent dynamic matching.

[0136] 402. Based on the spatial vector distribution of pollutant diffusion direction in the pollution concentration gradient map, calculate the reverse compensation intensity value between the deformation angle of the diversion structure and the spatial vector distribution in the diversion control parameter combination;

[0137] In step 402, the reverse compensation intensity value guides the spatial correction capability of the guide vane angle to the pollutant diffusion direction, quantified by the cosine value of the reverse angle between the guide direction and the diffusion direction. Spatial vector distribution refers to the characteristic of physical quantities being distributed in space according to a specific law. The deformation angle of the guide structure guides the adjustable angle of the guide structure in terms of physical shape, used to change the flow direction or velocity of the fluid.

[0138] In this embodiment, firstly, the spatial vector distribution of the pollutant diffusion direction is converted into an airflow direction vector representing the deformation angle of the guiding structure. Secondly, the combination of guiding control parameters is converted into a spatial vector. The angular deviation between the guiding angle and the diffusion direction is calculated using a cosine similarity algorithm, and a reverse compensation intensity value is generated by combining this with the projection length ratio. The reverse compensation intensity value is used to quantify the ability of the guiding structure to adjust the pollutant diffusion path; a higher value indicates that the guiding angle is more effective in counteracting the influence of the diffusion direction.

[0139] 403. Assign a spatial coverage priority coefficient to the combination of diversion control parameters according to the reverse compensation intensity value, and sort the combination of diversion control parameters based on the spatial coverage priority coefficient to select a subset of target parameters that cover the high concentration area in the pollution concentration gradient map.

[0140] In step 403, the spatial coverage priority coefficient refers to a comprehensive priority score calculated based on the reverse compensation intensity and the spatial coverage range of the pollution concentration gradient, used to screen the optimal parameter combination. A high-concentration area refers to a region within a specific spatial range where the pollutant concentration is significantly higher than the surrounding environment. The target parameter subset refers to the optimal or highly suitable set of parameters extracted from complex parameter combinations through a screening algorithm.

[0141] In this embodiment, firstly, the reverse compensation intensity value is combined with the coverage area of ​​high-concentration regions in the pollution concentration gradient map, and a spatial coverage priority coefficient is calculated using a weighted scoring algorithm. Then, the diversion control parameter combinations are sorted from high to low according to the spatial coverage priority coefficient, and parameter combinations that meet the coverage ratio of high-concentration regions are selected, ultimately forming a subset of target parameters. This ensures that both high correction capability and wide-area coverage requirements are simultaneously met.

[0142] 404. Generate a dynamic flow guidance parameter set based on the spatial synergistic effect strength between the flow guidance structure deformation angle and the air volume distribution parameter in the target parameter subset.

[0143] In step 404, the spatial synergy strength guides the synergistic blocking effect of the combination of the guide vane angle and airflow distribution parameters on the pollutant diffusion path, which is quantified by the correlation between the airflow disturbance range and the concentration decrease rate.

[0144] In this embodiment, firstly, the deformation angle of the guide structure is weighted by priority coefficients to generate a comprehensive guide angle. Secondly, a fuzzy clustering algorithm is used to merge similar strategies for the airflow distribution parameters to optimize the distribution ratio. Subsequently, the synergistic effect of the parameter combination is verified through an airflow field simulation model, and interference terms are eliminated. Finally, a dynamic guide parameter set containing the deformation angle of the guide structure, airflow distribution parameters, and disturbance frequency is output, adapting to the current pollution diffusion state.

[0145] Here is a specific example:

[0146] Within the subway station, pollutants primarily diffuse from the platform to the concourse, spreading in a northeast direction. First, all possible combinations of guiding structure deformation angles and airflow distribution parameters are extracted from the flow control characteristics. A clustering algorithm identifies candidate combinations with a high degree of matching to the northeast diffusion direction, and a Gaussian diffusion model is used to verify their pollutant interception efficiency. Finally, suitable parameter combinations are selected. Based on the spatial vector distribution of pollutant diffusion directions in the pollution concentration gradient map, the selected guiding angles are converted into spatial vectors. A cosine similarity algorithm is used to calculate the angle deviation between the selected angle and the diffusion direction, and a reverse compensation intensity value is generated by combining the projection length ratio to quantify the ability of the guiding structure to adjust the pollutant path. Furthermore, considering the coverage area of ​​high-concentration regions, a weighted scoring algorithm is used to calculate the spatial coverage priority coefficient, and the guiding parameter combinations are sorted in descending order of priority to select highly suitable parameters covering high-concentration areas. A spatial synergy analysis is performed on the selected target parameter subset. Monte Carlo simulation is used to evaluate the joint contribution of guiding angles and airflow parameters, ultimately generating a dynamic guiding parameter set. The guiding structure is adjusted in real time to intercept pollutants, achieving dynamic adaptation in pollution control.

[0147] In summary, steps 401 to 404, by associating the flow control features with the pollution concentration gradient map, achieved precise adaptation of the dynamic flow parameters. This ensures that the flow parameters match the pollutant diffusion direction and cover high-concentration areas. Furthermore, through reverse compensation and synergistic optimization, pollution control efficiency is significantly improved. In the subway station scenario, the dynamic flow parameter set can increase the pollutant interception rate and support adaptive adjustment under real-time environmental changes, providing an efficient and flexible technical solution for open-environment pollution control.

[0148] To address the joint analysis of multidimensional environmental data and pollutant distribution heatmaps, this scheme generates a pollution concentration gradient map by jointly encoding the multidimensional environmental data and the pollutant distribution heatmap. First, the dynamic correlation between the migration direction of pollutant particles and real-time airflow velocity is calculated. Based on this relationship, the diffusion path of pollutants in three-dimensional space is predicted, and a real-time interactive mapping relationship is established in conjunction with the concentration distribution. Finally, the static pollution heatmap is superimposed with the dynamic airflow velocity to generate a pollution concentration gradient map with temporal continuity, providing dynamic input for subsequent control. In some embodiments, step 102, which involves jointly encoding the multidimensional environmental data and the pollutant distribution heatmap to generate the pollution concentration gradient map, includes:

[0149] 501. Based on the pollution concentration gradient of the pollutants in the pollutant distribution heat map, calculate the migration direction of the pollutant particles, and based on the pollution concentration values ​​at each spatial location in the pollutant distribution heat map, determine the dynamic correlation between the migration direction of the pollutant particles and the real-time airflow velocity in the multidimensional environmental data.

[0150] In step 501, the direction of pollutant particle migration refers to the spatial movement trend of pollutants under the influence of a concentration gradient, represented by a vector direction from a high-concentration area to a low-concentration area. The dynamic correlation refers to the interaction between the direction of pollutant particle migration and the real-time airflow velocity, reflecting the driving effect of airflow on the pollutant diffusion path. The pollutant concentration gradient refers to the concentration change trend of pollutants in space due to uneven distribution, typically manifested as diffusion from high-concentration areas to low-concentration areas. The real-time airflow velocity refers to the speed and direction of airflow in the current environment.

[0151] In this embodiment, firstly, based on the pollution concentration gradient of the pollutant distribution heatmap, the pollution concentration value at each spatial location is calculated using the finite difference method or image gradient algorithm to determine the migration direction of pollutant particles. Then, combining the vector information of real-time airflow velocity in multidimensional environmental data, a vector projection algorithm is used to compare the migration direction of pollutant particles with the airflow velocity direction, calculating the angle deviation and projection length ratio to quantify the dynamic correlation between the two. Finally, the above correlation parameters are input into a dynamic regression model to train and obtain the response function of the pollution particle migration direction to changes in airflow velocity, thus constructing the dynamic correlation.

[0152] 502. Based on the dynamic correlation, calculate the predicted diffusion path of the pollutant in three-dimensional space affected by the real-time airflow velocity, and based on the spatial distribution relationship between the predicted diffusion path and the pollutant concentration value, establish a real-time interactive mapping relationship between the pollutant concentration gradient and the real-time airflow velocity.

[0153] In step 502, the diffusion path prediction result refers to the migration trajectory of pollutants in three-dimensional space simulated based on dynamic correlations. Three-dimensional space refers to the objectively existing space composed of length, width, and height, which is the physical space perceived by humans in daily life. Spatial distribution relationships refer to the relative positions, arrangement order, and interaction patterns of geographical entities or phenomena in space, typically described through topological, ordinal, and metric relationships. Real-time interaction mapping relationships refer to the dynamic coupling model of pollution concentration gradient and real-time airflow velocity, used to describe the mathematical relationship of their real-time interactive influence.

[0154] In this embodiment, firstly, based on the migration direction correction coefficient of the dynamic correlation and the real-time airflow velocity vector, the pollutant particles are discretized into a set of point masses using the Lagrange particle tracking method, and assigned initial positions and adjusted velocity parameters. The displacement trajectory of the point masses in three-dimensional space is simulated using the Euler integral method to generate diffusion path prediction results. Subsequently, using a Gaussian diffusion model combined with the spatial coordinates and timestamps of the point mass trajectories, the distribution function of pollutant concentration values ​​changing over time is calculated, and the accuracy of the model is verified through Monte Carlo simulation. Finally, a real-time interactive mapping relationship between the pollutant concentration gradient and the real-time airflow velocity is established.

[0155] 503. Through the real-time interactive mapping relationship, the static concentration distribution of the pollutant distribution heat map is dynamically superimposed with the real-time airflow velocity to generate a pollutant concentration gradient map with time continuity.

[0156] In step 503, static concentration distribution refers to the stable concentration distribution of pollutants or other substances in space at a fixed point in time or under conditions of no external dynamic disturbance. Dynamic overlay refers to the dynamic fusion of data or phenomena from different time or spatial dimensions in a real-time changing environment to generate a comprehensive result reflecting the interaction relationships. Temporal continuity means that the map can reflect the dynamic change process of pollutant concentration over time, rather than a static snapshot at a single moment.

[0157] In this embodiment, firstly, the diffusion path prediction results are combined with the static concentration values ​​of the pollutant distribution heatmap. A spatiotemporal interpolation algorithm is then used to spatially-temporally fuse the static concentration distribution with real-time airflow velocity. Airflow velocity parameters are updated with a fixed step size on the time axis, and the spatial distribution of pollutant concentration is adjusted according to the airflow velocity direction on the spatial axis. Finally, the superimposed data is input into a dynamic visualization engine to generate a temporally continuous pollutant concentration gradient map.

[0158] Here is a specific example:

[0159] In a subway station, respirable particulate matter diffuses from the platform to the concourse, spreading northeastward, with high concentrations concentrated in the northwest corner. A heat map of pollution distribution is generated using a sensor network deployed in the concourse, and the pollutant concentration gradient in this heat map is used to calculate that the migration direction of pollutant particles in the northwest corner of the concourse is northeastward. Combining real-time airflow velocity monitoring by wind speed sensors, a vector projection algorithm is used to quantify the angular deviation and projection length ratio between the pollution migration direction and airflow velocity, thus establishing a dynamic correlation model between the pollution migration direction and changes in airflow velocity. Subsequently, the Lagrange particle tracking method is used to simulate the diffusion path of respirable particulate matter from the platform to the northwest corner of the concourse, predicting that pollutants will cover the northwest corner of the concourse in a short period. The concentration distribution is calculated using a Gaussian diffusion model, revealing that as wind speed increases, the pollutant peak shifts northeastward. Finally, the static pollution distribution heat map and real-time wind speed data are fused together using a spatiotemporal interpolation algorithm to generate a time-continuous concentration gradient map of inhalable particulate matter. This map visually presents the dynamic process of pollutants spreading from the platform to the northwest corner of the station hall in a time-series animation, and the diffusion range is adjusted in real time according to changes in wind speed.

[0160] In summary, steps 501 to 503, through multi-source data fusion and dynamic modeling, achieve deep interaction between pollutant distribution heatmaps and real-time environmental data, significantly improving the intelligence level of pollution control. Specifically, the pollution concentration gradient map has dynamic adaptability, responding in real time to changes in airflow velocity, thereby accurately predicting pollutant migration trajectories and optimizing control measures. Time-series animations visually present the spatiotemporal evolution of pollutant diffusion, providing decision-makers with dynamic control data. Furthermore, this solution is compatible with complex environmental scenarios, using heatmaps to locate high-concentration areas and generate problem lists, providing data support for targeted operations and task assignment, achieving precise and efficient pollution control.

[0161] To address the issues of dynamic airflow interference and uneven concentration distribution in pollutant diffusion path prediction, the scheme generates prediction results through vector superposition and concentration correction. First, the pollutant particle migration direction vector and the real-time airflow velocity vector are superimposed in three dimensions to generate an initial diffusion path prediction vector set. Then, the vector directions are corrected based on the pollutant concentration gradient to eliminate interference from low-concentration areas, ultimately obtaining a path prediction result that conforms to actual diffusion patterns, providing crucial input for the dynamic updating of the pollutant concentration gradient map. In some embodiments, step 502, which involves calculating the diffusion path prediction result of the pollutant in three-dimensional space affected by the real-time airflow velocity based on the dynamic correlation, includes:

[0162] 601. Based on the dynamic correlation, the spatial vector of the migration direction of the pollutant particles is superimposed with the vector of the real-time airflow velocity in three-dimensional space to generate the initial diffusion path prediction vector set of the pollutants;

[0163] In step 601, a spatial vector refers to a geometric object with magnitude and direction in three-dimensional space, typically used to describe the spatial distribution of physical quantities or the motion characteristics of an object. The initial diffusion path prediction vector set refers to the preliminary diffusion path set generated by superimposing the pollutant particle migration direction vector and the real-time airflow velocity vector, reflecting the initial driving effect of airflow on pollutant migration. Three-dimensional spatial superposition refers to the vector synthesis of the pollutant particle migration direction and real-time airflow velocity vectors in a three-dimensional coordinate system.

[0164] In this embodiment, firstly, a spatial vector representing the direction of pollution migration is extracted from pollution source monitoring data. Simultaneously, a real-time airflow velocity vector, including horizontal and vertical wind speed components, is acquired using a wind speed sensor. Secondly, based on the principle of vector superposition, the components of the pollution migration vector and the airflow velocity vector are added to generate a composite vector. Finally, the trajectory of the composite vector is simulated using the Lagrange particle tracking method. By iteratively calculating the motion path of pollution particles in three-dimensional space, an initial diffusion path prediction vector set is generated.

[0165] 602. Based on the pollution concentration values ​​at each spatial location in the pollutant distribution heatmap, the direction of each vector in the initial diffusion path prediction vector set is corrected by concentration gradient to obtain the diffusion path prediction result.

[0166] In step 602, each spatial location refers to various classifications and definitions describing the positional relationships of objects, regions, or phenomena in three-dimensional space in different fields or application scenarios. Concentration gradient correction refers to adjusting the direction and rate of the initial diffusion vector based on the concentration values ​​at each location in the pollutant distribution heatmap, so that the predicted path better matches the actual concentration decay law.

[0167] In this embodiment, firstly, the pollution concentration values ​​at each spatial location are extracted from the pollutant distribution heatmap, and the concentration gradient direction is calculated using the finite difference method. Secondly, the directions of each vector in the initial diffusion path prediction vector set are compared with the concentration gradient direction to identify the deviation angle between the two, which serves as the basis for correction. Subsequently, a vector correction algorithm is used to weightedly fuse the initial vector direction and the concentration gradient direction, adjusting the migration direction of pollutant particles to better reflect the actual concentration distribution characteristics. Finally, the corrected vectors are substituted into the Lagrange particle tracking model to re-simulate the trajectory of pollutant particles, generating the corrected diffusion path prediction results.

[0168] Here is a specific example:

[0169] After deploying a network of inhalable particulate matter sensors and wind speed sensors in the subway station concourse, monitoring revealed that the pollutant concentration gradient spreading from the platform to the concourse was oriented northeastward, based on real-time airflow velocity. First, the pollution migration vector and airflow velocity vector were superimposed in three-dimensional space to generate a composite vector. Then, using the Lagrange particle tracking method, the trajectory of pollutant particles was simulated, predicting that pollutants would cover the northwest corner of the concourse in ten minutes. By calculating the deviation between the concentration gradient direction in the northwest corner of the concourse and the initial diffusion direction, the particle movement direction was corrected. The trajectory showed that the pollutant peak shifted northeastward, dynamically adjusted in conjunction with the influence of wind speed. The resulting continuous time-series concentration gradient map visually presented the changes in the pollutant diffusion range and concentration distribution, assisting decision-makers in effectively intercepting pollutants by adjusting the angle of the deflector, increasing the tilt of the northeast deflector, thus achieving precise and dynamic pollution control.

[0170] In summary, steps 601 to 602, through a two-stage processing approach of three-dimensional spatial vector superposition and concentration gradient correction, achieve dynamic and accurate prediction of pollutant diffusion paths. The vector superposition principle is used to fuse pollutant migration with airflow velocity to generate an initial diffusion path. The path direction is further corrected based on the concentration gradient method, resolving the coupling problem between airflow and uneven concentration distribution. In a subway station scenario, this method can dynamically respond to changes in wind speed and, combined with thermal mapping, locate high-concentration areas, significantly improving the accuracy and timeliness of pollution control.

[0171] To further improve the dynamic response capability of air purification equipment in complex polluted environments, the solution generates control commands through dynamic weight allocation and boundary correction. First, execution weights are dynamically allocated based on the distribution density and diffusion rate of high-concentration areas, generating a preliminary set of execution parameters. Then, boundary constraints are corrected by considering the physical deformation limitations of the airflow guiding structure. Finally, the corrected airflow angle and airflow parameters are matched and combined according to spatial coverage density to generate executable intelligent control commands, ensuring the safety of equipment operation and the accuracy of pollution control. In some embodiments, step 105, generating the intelligent control commands for the air purification equipment based on the dynamic airflow parameter set, includes:

[0172] 701. Based on the spatial distribution density of high-concentration areas in the pollution concentration gradient map and the real-time rate of change of pollutant diffusion direction, assign dynamic execution weight values ​​to the dynamic diversion parameter set to generate the execution parameter set;

[0173] In step 701, the dynamic execution weight value refers to the priority score assigned to the flow guidance parameters based on the spatial distribution density of the high-concentration area and the rate of change of the pollutant diffusion direction. The execution parameter set refers to the set of parameters including the flow guidance angle, air volume allocation ratio, and corresponding weight values.

[0174] In this embodiment, firstly, a density-based spatial clustering algorithm is used to analyze the spatial distribution density of high-concentration areas in the pollution concentration gradient map to identify the pollutant aggregation intensity. Secondly, the trajectory of pollutant particles is simulated using the Lagrange particle tracking method, and the real-time rate of change of pollutant diffusion direction is calculated using the time series difference method to quantify the dynamic trend of pollutant migration. Subsequently, a multi-objective optimization algorithm is used with spatial distribution density and diffusion rate as input parameters to calculate the dynamic execution weight value corresponding to each region, prioritizing the control of high-risk areas. Finally, the dynamic weight values ​​are mapped to a dynamic diversion parameter set to generate an execution parameter set.

[0175] 702. Based on the physical deformation range limitation of the adjustable airflow guiding structure of the air purification device, the deformation angle of the airflow guiding structure in the set of execution parameters is corrected by boundary constraints to obtain the corrected deformation angle of the airflow guiding structure.

[0176] In step 702, the air purification equipment refers to a device that removes suspended particles, harmful gases, and odors from the air through technologies such as physical filtration, chemical adsorption, and ionization sterilization. The physical deformation range limitation refers to the mechanically adjustable range of the air purification equipment's airflow guiding structure. An adjustable airflow guiding structure refers to a device that dynamically changes the direction and distribution of airflow through mechanical deformation or position adjustment. Boundary constraint correction refers to adjusting the airflow guiding angle parameters according to the equipment's physical limitations to ensure that commands are executable without damaging the equipment. The corrected airflow guiding structure deformation angle refers to the final value after boundary constraint correction of the initially calculated deformation angle within the physical deformation limit range.

[0177] In this embodiment, firstly, a physical constraint model for the deformation angle of the adjustable airflow guiding structure is established according to the design specifications. Secondly, the execution parameter set is input into the constraint verification module to check whether the deformation angle of the airflow guiding structure exceeds the physical deformation range. Subsequently, mathematical constraint correction technology is used to trim or optimize the deformation angle values ​​of the airflow guiding structure that exceed the limits, ensuring that they conform to the characteristics of the air purification equipment. Finally, the corrected deformation angle of the airflow guiding structure is output, ensuring the safety and feasibility of the airflow guiding structure's operation.

[0178] 703. The modified airflow guiding structure deformation angle and the corresponding airflow distribution parameters are matched and combined according to the spatial coverage density of the dynamic airflow guiding parameter set to generate intelligent control commands.

[0179] In step 703, spatial coverage density refers to the proportion of high-concentration areas effectively covered by the guide flow parameter combination in the pollution concentration gradient map. Matching combination refers to binding the corrected guide flow angle with the airflow distribution parameters according to the optimal coverage density principle.

[0180] In this embodiment, firstly, the air purification area is divided into a three-dimensional grid, and the pollution coverage density of each grid is calculated. Secondly, a multi-dimensional matching algorithm is used to associate the corrected airflow guiding structure deformation angle with the airflow distribution parameters according to the spatial coverage density; for example, a higher density area is matched with a larger deformation angle and a higher airflow. Subsequently, the matching results are converted into specific instructions and output in a protocol format that the device can parse. In this process, spatial coverage density serves as the matching basis, driving the coordinated optimization of deformation angle and airflow parameters, ultimately generating executable intelligent control instructions to achieve precise control of the air purification equipment.

[0181] Here is a specific example:

[0182] In the subway station concourse, sensor networks detected a pollutant concentration gradient spreading from the platform to the concourse in the northeast direction, with high-concentration areas concentrated in the northwest corner. Pollutant concentration gradient map analysis showed that the northwest corner was a high-density area, and the real-time rate of change of its pollutant diffusion direction was analyzed. A dynamic weight allocation algorithm was used, with the spatial distribution density and diffusion rate of the high-concentration area as input parameters, to assign dynamic execution weight values ​​to the northwest corner and other areas, generating a preliminary execution parameter set. Subsequently, based on the physical deformation range limitations of the guide vane, boundary constraints were applied to the deformation angles in the execution parameter set. Confirming that the limits were not exceeded, these angles were retained as the corrected guide vane deformation parameters. Finally, combining the spatial coverage density of the dynamic guide vane parameter set, the corrected deformation angles and airflow parameters were matched and combined according to the coverage characteristics of the high-density area to generate intelligent control commands. This drove the guide vane to adjust towards the target direction and enhance airflow, achieving precise interception and dilution of high-concentration pollutants in the northwest corner.

[0183] In summary, steps 701 to 703, through a combination of dynamic weight allocation, physical constraint correction, and spatial matching, achieve a precise response of the air purification equipment to pollution diffusion. In the subway station scenario, the synergistic optimization of the deflector angle adjustment and airflow distribution strategy can reduce the pollutant concentration in the northwest corner within ten minutes, while simultaneously reducing equipment energy consumption. Compared to traditional fixed control methods, this approach significantly improves pollution interception efficiency and dynamic adaptability, providing an intelligent solution for air quality control in complex environments.

[0184] Figure 2 This application provides a schematic diagram of the structure of an intelligent control system for an air purification device, as shown in the embodiment. Figure 2 As shown, the system includes:

[0185] The acquisition module 21 acquires multidimensional environmental data of pollutants in an open polluted environment. The multidimensional environmental data includes three-dimensional spatial data of the pollutants, real-time airflow velocity, and pollutant concentration distribution data captured by the air purification equipment.

[0186] Encoding module 22 generates a pollutant distribution heat map based on the three-dimensional spatial data, and jointly encodes the multi-dimensional environmental data with the pollutant distribution heat map to generate a pollution concentration gradient map;

[0187] Matching module 23 synchronously collects control strategy datasets of historical polluted environments, and performs cross-environment matching between the control strategy datasets and the pollution concentration gradient map to extract diversion control features that are spatially correlated with the pollutant diffusion direction in the pollution concentration gradient map;

[0188] The mapping module 24 determines the dynamic flow guidance parameter set in the open pollution environment based on the correlation mapping relationship between the flow guidance control features and the pollution concentration gradient map.

[0189] The generation module 25 generates intelligent control instructions for the air purification device based on the dynamic flow guidance parameter set. The intelligent control instructions include a directional airflow disturbance strategy and an airflow distribution scheme for the high-concentration area in the pollution concentration gradient map.

[0190] Figure 2 The intelligent control system of the air purification device described above can execute... Figure 1 The implementation principle and technical effects of the intelligent control method for an air purification device described in the illustrated embodiment will not be repeated here. The specific methods by which each module and unit performs operations in the intelligent control system of the air purification device in the above embodiments have been described in detail in the embodiments related to this method, and will not be elaborated upon here.

[0191] In one possible design, Figure 2 The intelligent control system of an air purification device in the illustrated embodiment can be implemented as a computing device, such as... Figure 3 As shown, the computing device may include a storage component 31 and a processing component 32;

[0192] The storage component 31 stores one or more computer instructions, wherein the one or more computer instructions are invoked and executed by the processing component 32.

[0193] The processing component 32 is used for the above Figure 1 The embodiment describes an intelligent control method for an air purification device.

[0194] The processing component 32 may include one or more processors to execute computer instructions to complete all or part of the steps in the above-described method. Alternatively, the processing component may be implemented as one or more application-specific integrated circuits (ASICs), digital signal processors (DSPs), digital signal processing devices (DSPDs), programmable logic devices (PLDs), field-programmable gate arrays (FPGAs), controllers, microcontrollers, microprocessors, or other electronic components to perform the above-described method.

[0195] Storage component 31 is configured to store various types of data to support operations at the terminal. The storage component can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic storage, flash memory, magnetic disk, or optical disk.

[0196] Of course, computing devices may also include other components, such as input / output interfaces, display components, communication components, etc.

[0197] Input / output interfaces provide interfaces between processing components and peripheral interface modules, which can be output devices, input devices, etc.

[0198] The communication components are configured to facilitate wired or wireless communication between computing devices and other devices.

[0199] The computing device can be a physical device or an elastic computing host provided by a cloud computing platform. In this case, the computing device can refer to a cloud server, and the aforementioned processing components, storage components, etc., can be basic server resources rented or purchased from the cloud computing platform.

[0200] This application also provides a computer storage medium storing a computer program, which, when executed by a computer, can perform the above-described functions. Figure 1 The embodiment shown illustrates an intelligent control method for an air purification device.

[0201] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working processes of the systems, devices, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0202] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs. Those skilled in the art can understand and implement this without any creative effort.

[0203] Through the above description of the embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus necessary general-purpose hardware platforms, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solutions, in essence or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods described in the various embodiments or some parts of the embodiments.

[0204] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit them. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of this application.

Claims

1. A method of intelligent control of an air purification apparatus, characterized by, The method comprises the following steps: acquiring multi-dimensional environmental data of pollutants in an open pollution environment, the multi-dimensional environmental data comprising three-dimensional spatial data of the pollutants captured by an air purification device, real-time airflow velocity, and pollutant concentration distribution data; generating a pollutant distribution heat map based on the three-dimensional spatial data, and jointly encoding the multi-dimensional environmental data and the pollutant distribution heat map to generate a pollution concentration gradient atlas; synchronously collecting a control strategy data set of a historical pollution environment, and performing cross-environment matching between the control strategy data set and the pollution concentration gradient atlas to extract a flow guiding control feature that is spatially correlated with a pollutant diffusion direction in the pollution concentration gradient atlas; determining a dynamic flow guiding parameter set in the open pollution environment according to the correlation mapping relationship between the flow guiding control feature and the pollution concentration gradient atlas; generating an intelligent control instruction of the air purification device according to the dynamic flow guiding parameter set, the intelligent control instruction comprising a directional airflow disturbance strategy and a wind volume allocation scheme for a high-concentration region in the pollution concentration gradient atlas; wherein the cross-environment matching between the control strategy data set and the pollution concentration gradient atlas to extract the flow guiding control feature that is spatially correlated with the pollutant diffusion direction in the pollution concentration gradient atlas comprises: filtering a candidate control strategy subset with a similarity to a spatial vector distribution of the pollutant diffusion direction reaching a preset threshold from the control strategy data set based on the spatial vector distribution in the pollution concentration gradient atlas; performing spatial correlation analysis on a flow guiding structure deformation angle and a wind volume allocation parameter in the candidate control strategy subset to establish a dynamic matching relationship between flow guiding control parameters in the control strategy data set and the pollutant diffusion direction; extracting, from the candidate control strategy subset, a flow guiding control parameter combination whose spatial vector included angle of the pollutant diffusion direction meets a preset condition according to the dynamic matching relationship, and calculating a dynamic priority weight value of each flow guiding control parameter combination based on a spatial coverage range of the flow guiding control parameter combination in the pollution concentration gradient atlas; fusing and reconstructing the flow guiding control parameter combination according to the dynamic priority weight value to generate the flow guiding control feature that is spatially correlated with the pollutant diffusion direction; the jointly encoding the multi-dimensional environmental data and the pollutant distribution heat map to generate a pollution concentration gradient atlas comprises: calculating a pollution particle migration direction of the pollutant according to a pollution concentration gradient of the pollutant in the pollutant distribution heat map, and determining a dynamic correlation relationship between the pollution particle migration direction and a real-time airflow velocity in the multi-dimensional environmental data based on a pollution concentration value of each spatial position in the pollutant distribution heat map; calculating a diffusion path prediction result of the pollutant in a three-dimensional space affected by the real-time airflow velocity according to the dynamic correlation relationship, and establishing a real-time interaction mapping relationship between the pollution concentration gradient and the real-time airflow velocity based on a spatial distribution relationship between the diffusion path prediction result and the pollution concentration value. The static concentration distribution of the pollutant distribution thermodynamic map is dynamically superimposed with the real-time airflow velocity through the real-time interactive mapping relationship, to generate a pollutant concentration gradient atlas with time continuity; The diffusion path prediction result of the pollutant in the three-dimensional space affected by the real-time airflow velocity is calculated according to the dynamic correlation relationship, including: According to the dynamic correlation relationship, the spatial vector of the migration direction of the pollution particle is superimposed with the vector of the real-time airflow velocity in three-dimensional space to generate an initial diffusion path prediction vector set of the pollutant; Based on the pollution concentration value of each spatial position in the pollutant distribution thermodynamic map, the direction of each vector in the initial diffusion path prediction vector set is corrected by the concentration gradient to obtain the diffusion path prediction result.

2. The method of claim 1, wherein, The spatial correlation analysis of the guide flow structure deformation angle and the air volume distribution parameter in the candidate control strategy subset is performed to establish the dynamic matching relationship between the guide flow control parameter in the control strategy data set and the diffusion direction of the pollutant, including: The guide flow structure deformation angle in the candidate control strategy subset is projected and compared with the diffusion direction of the pollutant in space to generate a first geometric correlation feature; The air volume distribution parameter in the candidate control strategy subset is regionally density-matched with the spatial coverage range of the corresponding pollution concentration gradient to generate a second geometric correlation feature; Based on the superimposition relationship between the first geometric correlation feature and the second geometric correlation feature, a dynamic matching rule library of the guide flow control parameter in the control strategy data set and the diffusion direction of the pollutant is established; According to the spatial vector distribution of the diffusion direction of the pollutant in the pollution concentration gradient atlas, a mapping relationship set that meets a preset condition is screened from the dynamic matching rule library to obtain a screened mapping relationship set; According to the screened mapping relationship set, the guide flow structure deformation angle and the air volume distribution parameter are bound according to the spatial correlation intensity to generate the dynamic matching relationship between the guide flow control parameter and the diffusion direction of the pollutant.

3. The method of claim 1, wherein, The dynamic guide flow parameter set in the open pollution environment is determined according to the correlation mapping relationship between the guide flow control feature and the pollution concentration gradient atlas, including: A guide flow control parameter combination associated with the diffusion direction of the pollutant in the pollution concentration gradient atlas is extracted from the guide flow control feature; Based on the spatial vector distribution of the diffusion direction of the pollutant in the pollution concentration gradient atlas, the reverse compensation intensity value of the guide flow structure deformation angle in the guide flow control parameter combination and the spatial vector distribution is calculated; According to the reverse compensation intensity value, a spatial coverage priority coefficient is assigned to the guide flow control parameter combination, and the guide flow control parameter combination is sorted based on the spatial coverage priority coefficient to screen a target parameter subset covering a high-concentration area in the pollution concentration gradient atlas; A dynamic guide flow parameter set is generated according to the spatial synergistic effect intensity of the guide flow structure deformation angle and the air volume distribution parameter in the target parameter subset.

4. The method of claim 1, wherein, The intelligent control instruction of the air purification equipment is generated according to the dynamic guide flow parameter set, including: According to the spatial distribution density of the high concentration area in the pollution concentration gradient map and the real-time change rate of the pollution diffusion direction, a dynamic execution weight value is assigned to the dynamic guide flow parameter set, and an execution parameter set is generated; Based on the physical deformation range limitation of the adjustable guide flow structure of the air purification equipment, the guide flow structure deformation angle in the execution parameter set is boundary-constrained and corrected, and a corrected guide flow structure deformation angle is obtained; The corrected guide flow structure deformation angle and the corresponding air volume allocation parameter are matched and combined according to the spatial coverage density of the dynamic guide flow parameter set, and an intelligent control instruction is generated.

5. The intelligent control system of an air purification device, applied to the intelligent control method of an air purification device according to any one of claims 1-4, characterized in that, Comprise: An acquisition module acquires multi-dimensional environmental data of pollutants in an open pollution environment, the multi-dimensional environmental data including three-dimensional spatial data of the pollutants captured by an air purification equipment, real-time air flow velocity, and pollution concentration distribution data; An encoding module generates a pollution distribution heat map based on the three-dimensional spatial data, and jointly encodes the multi-dimensional environmental data and the pollution distribution heat map to generate a pollution concentration gradient map; A matching module synchronously acquires a control strategy data set of a historical pollution environment, and cross-environmentally matches the control strategy data set and the pollution concentration gradient map to extract a guide flow control feature that is spatially correlated with the pollution diffusion direction in the pollution concentration gradient map; A mapping module determines a dynamic guide flow parameter set in the open pollution environment according to the correlation mapping relationship between the guide flow control feature and the pollution concentration gradient map; A generation module generates an intelligent control instruction of the air purification equipment according to the dynamic guide flow parameter set, the intelligent control instruction including a directional air flow disturbance strategy and an air volume allocation scheme for a high concentration area in the pollution concentration gradient map.

6. A computing device, comprising: A processing component and a storage component are included; the storage component stores one or more computer instructions; the one or more computer instructions are used to be called and executed by the processing component to realize the intelligent control method of the air purification equipment according to any one of claims 1-4.

7. A computer storage medium, characterized in that A computer program is stored, and when the computer program is executed by a computer, the intelligent control method of the air purification equipment according to any one of claims 1-4 is realized.

Images