Air quality sensor based on internet of things and remote control method thereof

Abstract

This invention relates to the field of air quality monitoring technology, and provides an Internet of Things (IoT)-based air quality sensor and its remote control method. The air quality sensor includes: a multi-source heterogeneous sensing fusion system, a spatiotemporal edge gateway system, and a dynamic response collaborative central system. The remote control method includes: processing the purification device linkage control matrix through heterogeneous command fusion technology to obtain a spatially optimized control vector set, which forms a human-machine interaction entropy value mechanism; processing the human-machine interaction entropy value model through an entropy weight dynamic allocation strategy to obtain a spatial state perception vector, which forms a cross-media alarm triggering channel; processing the cross-media alarm triggering channel through a bidirectional response coupling engine to obtain a device state feature topology, which forms a closed-loop control entropy correction command. This invention dynamically transforms pollution in physical space into operable device control parameters, establishing a complete IoT closed-loop system from environmental monitoring to purification execution.

Description

Technical Field

[0001] This invention relates to the field of air quality monitoring technology, and in particular to an air quality sensor based on the Internet of Things and its remote control method. Background Technology

[0002] Driven by the wave of smart city construction, the global IoT air quality monitoring market is projected to exceed $12 billion by 2025. Currently, while approximately 18,000 monitoring stations have been deployed in key areas of my country, traditional equipment suffers from three major drawbacks: limited coverage radius per station, making it impossible to build high-resolution pollution maps; long data transmission delays, hindering emergency decision-making; and high maintenance costs, accounting for a significant portion of total investment.

[0003] Existing technology 1, application number: CN 202410521776.1, discloses an intelligent air quality monitoring system and method based on Internet of Things (IoT) technology. It uses air quality sensors to monitor and collect air parameters in real time, such as temperature, humidity, carbon dioxide concentration, sulfur dioxide concentration, ozone concentration, and particulate matter concentration. The data is then transmitted to a backend server using IoT technology. The backend server employs AI-based data processing and analysis algorithms to perform time-series collaborative correlation analysis of these air parameter data, thereby detecting the air quality level. While this approach can improve the accuracy of air quality level judgment based on the dynamic and coordinated changes of various air components, thus providing timely and accurate air quality information and a scientific basis for air pollution control and treatment, it lacks edge computing capabilities. This prevents the rapid extraction of pollution characteristics locally, leading to data redundancy and transmission burden. Furthermore, it only provides air quality level judgment and does not form a closed-loop control system from monitoring to purification, failing to directly link purification equipment for dynamic response.

[0004] Prior art 2, application number: CN202311092151.X, discloses a cloud-based environmental monitoring method and device, including: sensor selection: selecting sensors according to the environmental parameters to be monitored, including temperature sensors, humidity sensors, and air quality sensors; connecting sensors to IoT devices: connecting sensors to IoT devices, including IoT gateways or embedded devices; data acquisition and transmission: the IoT devices transmit the data collected by the sensors to the cloud platform via wireless communication, including wireless networks, cellular networks, or LoRaWAN; cloud data storage: the cloud platform receives and stores the sensor data; data processing and analysis: the cloud platform processes and analyzes the received environmental data. Although the cloud platform sends alarm signals in abnormal situations based on the processing and analysis results, it relies on cloud-based data storage and analysis without edge computing optimization, resulting in large data volumes, high transmission costs, and delays in cloud-based decision-making; it only performs basic data acquisition and alarms, without modeling the spatiotemporal migration patterns of pollutants, making it impossible to predict pollutant diffusion trends; the alarm mechanism only remains at the notification level and does not form an automated control matrix with purification equipment, failing to achieve precise local environmental optimization.

[0005] Current technologies 1 and 2 suffer from high latency in air quality monitoring and response, and their pollution prediction capabilities need further improvement. Therefore, this invention provides an Internet of Things (IoT)-based air quality sensor and its remote control method. Summary of the Invention

[0006] To address the aforementioned technical problems, this invention provides an air quality sensor based on the Internet of Things, comprising:

[0007] The multi-source heterogeneous sensing fusion system is used to form a three-dimensional sensing layer through spatial grid deployment of a monitoring network composed of multiple heterogeneous sensor nodes. Each heterogeneous sensor node collects the gaseous molecule concentration fluctuations in the physical space in real time, and obtains composite pollution characteristic factors through multi-source heterogeneous fusion processing. The composite pollution characteristic factors form a microenvironment quality evolution map.

[0008] The spatiotemporal edge gateway system is used to deploy computing node clusters on the edge side of the air quality monitoring area. It adopts a hierarchical aggregation architecture to process the sensing data. Each edge node performs edge topology compression processing on the micro-environment map within its coverage area to obtain spatiotemporal feature slices. The spatiotemporal feature slices form a confined space pollution migration cloud strategy.

[0009] The dynamic response coordination hub system is used to process the cloud strategy for migrating pollution in a confined space through a cross-domain strategy decoupling engine to obtain a set of device control vectors, which in turn form a linkage control matrix for the purification device.

[0010] Optional, a spatiotemporal domain edge gateway system, including:

[0011] The spatiotemporal dimension decomposition subsystem is used to extract the pollution propagation potential energy field and real-time pollution intensity scalar field from the microenvironment quality evolution map through four-dimensional tensor slicing, and generate a dual-channel spatiotemporal data stream.

[0012] The edge topology compression processing subsystem is used for processing dual-channel spatiotemporal data streams through an entropy-constrained collapse engine; it performs adaptive eddy detection on the pollution propagation potential energy field to capture the pollution migration direction vector; it performs spatial pyramid downsampling on the real-time pollution intensity scalar field to retain intensity gradient abrupt change points; and it generates high-information-density spatiotemporal feature slices by fusing the pollution migration direction vector and the intensity gradient abrupt change points through tensor cross product.

[0013] The pollution migration cloud strategy construction subsystem is used to calculate the probability distribution of pollution sources through nonlinear diffusion inversion using spatiotemporal feature slices as input units; generate equipotential surfaces of pollution concentration and tangent bundles of migration paths based on the probability distribution of pollution sources; and encode the equipotential surfaces and tangent bundles of migration paths into a dynamic strategy topology network to form a closed-space pollution migration cloud strategy that can describe the law of pollution diffusion.

[0014] Optional, an edge topology compression processing subsystem, comprising:

[0015] The pollution propagation potential energy field processing component is used to scan the potential energy gradient distribution of the pollution propagation potential energy field through a variable-scale differential operator, extract the curl extrema of the potential energy change, and generate a pollution migration direction vector set containing scalars of direction angle and vortex intensity. The pollution migration direction vector set is processed by a vector field sparsification engine to remove noise vectors with vortex intensity below the dynamic threshold, and to perform curvature consistency clustering on adjacent vectors. The output is a topology-simplified pollution migration main direction vector bundle.

[0016] The pollution intensity scalar field processing component is used to construct a second-order differential curvature map of the intensity field using a real-time pollution intensity scalar field, identify zero-crossing points in the curvature map as gradient abrupt change critical points, and establish a non-uniform sampling grid centered on the critical points. Through adaptive pyramid downsampling constrained by the critical points, intensity mean aggregation is performed in non-abrupt regions, and the original intensity values ​​are preserved with the gradient abrupt change critical points as the core, generating a set of pollution intensity gradient abrupt change points with weight labels.

[0017] The dual-channel data fusion component is used to convert the main direction vector bundle of pollution migration into a direction tensor field, map the set of abrupt change points of pollution intensity gradient into an intensity weight tensor, perform positive cross-multiplication of the two tensors, and output spatiotemporal feature slices that fuse the pollution migration direction and intensity abrupt change features.

[0018] Optional, dual-channel data fusion component, including:

[0019] The Directional Vector Beam to Tensor Field Sub-component is used to convert the main directional vector beam of pollution migration into a unit directional basis by using the curvature consistency clustering center of the vector beam as the reference point and through tangent space mapping, thereby generating a set of directional basis beams.

[0020] The tensor field generation sub-component is used to convert the vortex intensity scalar of each directional basis of the directional basis set into directional weight coefficients, construct a three-dimensional directional tensor with the directional basis as the axis and the weight as the modulus, and output a pollution migration directional tensor field with spatial distribution properties.

[0021] The mutation point set to weighted tensor quantum component is used to establish a radial decay function centered on the critical point of the pollution intensity gradient mutation point set, calculate the intensity amplitude of the decay function based on the weight label of the point set, and generate a mutation kernel function distribution cloud; the vertices of the non-uniform sampling grid are used as tensor control points, and bicubic spline interpolation of the kernel function value is performed between the control points to output a pollution intensity weighted tensor field covering the entire space.

[0022] Optional, the mutation set to weighted tensor quantum component includes:

[0023] The function distribution cloud generation module is used to input the set of pollution intensity gradient abrupt change points into the radial decay function constructor, extract the spatial coordinates and weight label values ​​of each critical point, establish an exponential decay function with the critical point as the center, and dynamically adjust the function decay coefficient by the weight label value to generate a decay function unit with amplitude modulation.

[0024] The kernel function cloud synthesis module is used to superimpose the function units of all decay function units according to spatial coordinates, eliminate the interference effect in the overlapping area by amplitude normalization, and output a continuously distributed abrupt kernel function distribution cloud; the non-uniform sampling grid vertex input tensor control point binder reads the spatial coordinates of the grid vertex as the reference position of the control point, injects the kernel function amplitude at the corresponding position at the control point, and generates a tensor control point set with amplitude constraints;

[0025] The global tensor interpolation module is used to construct a three-dimensional parametric surface with control points as nodes, perform continuous interpolation of kernel function amplitude along the surface normal, ensure smooth gradient of the interpolated surface through curvature constraints, and output a pollution intensity weight tensor field covering the entire domain.

[0026] Optional, a dynamic response coordination hub system, including:

[0027] The migration cloud strategy decoupling subsystem is used to extract the pollution concentration equipotential surface features in the pollution migration cloud strategy, generate the pollution intensity gradient tensor, separate the curvature parameters of the tangent bundles of the migration path, generate the path direction admittance matrix, and output the decoupled strategy feature primitives; match the pollution intensity gradient tensor with the power response curve of the purification equipment, map the path direction admittance matrix to the equipment spatial orientation topology; and generate a dimensionally orthogonalized set of equipment control vectors.

[0028] The linkage control matrix construction subsystem is used to construct the equipment operation unitary space based on the equipment control vector, and to allocate the weight of the equipment control quantity in the unitary space according to the direction of the pollution intensity gradient; to establish the phase coordination relationship of the equipment control quantity according to the equipment spatial orientation topology, and to encode the weight allocation and phase coordination into a multi-dimensional control plane, and output the linkage control matrix of the purification device that can drive multiple terminals.

[0029] A dual-channel feedback triggering subsystem is used to synchronously activate the feedback arbitrator in conjunction with the control matrix.

[0030] Optionally, the dual-channel feedback triggering subsystem includes: a mobile terminal channel, which, when the pollution intensity weight in the linkage control matrix exceeds the dynamic threshold, extracts the spatial code of the exceeding area, synthesizes the spatial code and intensity weight into an over-threshold alarm vector, and pushes a gaseous parameter over-threshold alarm to the gateway; after receiving the gaseous parameter over-threshold alarm, the gateway, according to a preset program, links the playback, lighting, and ventilation systems, etc., and the devices display the content of the gaseous parameter over-threshold alarm according to a preset program; at the same time, the gateway pushes the alarm information to the APP through the alarm information push platform; and a physical indication channel, which, according to the phase coordination relationship of the linkage control matrix, analyzes the key pollution nodes in the device orientation topology and triggers the spectral warning pulse sequence of the corresponding location device.

[0031] Optional, a linkage control matrix construction subsystem, including:

[0032] The phase coordination relationship establishment component is used to determine the spatial orientation topology of the equipment, calculate the shortest connection path length between any two equipment, convert the path length into the action delay time base, and generate a spatiotemporal delay field between the equipment. The spatiotemporal delay field generates a diffused wavefront with the pollution source location as the wave source. Based on the delay field, the phase offset angle of the control action of each equipment is calculated, and the equipment control phase difference vector is output.

[0033] The control tensor synthesis component is used to map the pollution intensity gradient weights to the control magnitude matrix and convert the phase difference vector into a time-series coordination operator.

[0034] The multidimensional plane construction component is used to construct a complex control plane with the magnitude matrix as the real part and the timing operator as the imaginary part. The complex plane is extended along the device space dimension to form a three-dimensional control tensor. Orthogonal projection compression is performed on the tensor to generate a multidimensional control plane.

[0035] Optionally, fluctuations in gaseous molecule concentration include air quality parameters such as PM2.5, CO2, TVOC, temperature, and humidity.

[0036] This invention provides a remote control method for an air quality sensor based on the Internet of Things, comprising the following steps:

[0037] The purification device linkage control matrix is ​​processed by heterogeneous instruction fusion technology to obtain a spatially optimized control vector set, which forms a human-machine interaction entropy value mechanism.

[0038] The entropy value model of human-computer interaction is processed by the entropy weight dynamic allocation strategy to obtain the spatial state perception vector, and the spatial state perception vector forms a cross-media alarm triggering channel.

[0039] The cross-media alarm triggering channel is processed by the bidirectional response coupling engine to obtain the device status feature topology, which forms a closed-loop control entropy correction command.

[0040] This invention achieves a three-tiered technological closed loop in the field of IoT air quality monitoring: A multi-source heterogeneous sensing fusion system transforms discrete gaseous molecule concentration data into spatially correlated composite pollution characteristic factors, overcoming the limitations of traditional single-point sensor measurements and constructing a quantifiable microenvironment quality evolution map. A spatiotemporal edge system performs topological compression on high-dimensional map data, generating feature slices containing both temporal and spatial attributes, reducing the amount of original monitoring data while still fully characterizing the migration patterns of pollutants in enclosed spaces. A dynamic response collaborative central system decouples abstract migration strategies into executable device control vectors, ultimately forming a control matrix that enables the coordinated response of a group of purification devices, transforming traditionally independently operating purification equipment into a spatially strategic interconnected network.

[0041] Other features and advantages of the invention will be set forth in the following description, and will be apparent in part from the description, or may be learned by practicing the invention. The objects and other advantages of the invention may be realized and obtained by means of the structures particularly pointed out in the written description and the accompanying drawings.

[0042] The technical solution of the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. Attached Figure Description

[0043] The accompanying drawings are provided to further illustrate the invention and form part of the specification. They are used in conjunction with embodiments of the invention to explain the invention and do not constitute a limitation thereof. In the drawings:

[0044] Figure 1 This is a block diagram of the IoT-based air quality sensor in Embodiment 1 of the present invention;

[0045] Figure 2 This is a block diagram of the multi-source heterogeneous sensing fusion system in Embodiment 2 of the present invention;

[0046] Figure 3 This is a block diagram of the spatiotemporal domain edge gateway system in Embodiment 3 of the present invention;

[0047] Figure 4 This is a block diagram of the dynamic response coordination central system in Embodiment 7 of the present invention;

[0048] Figure 5 This is a flowchart of the remote control method for an air quality sensor based on the Internet of Things in Embodiment 11 of the present invention. Detailed Implementation

[0049] The preferred embodiments of the present invention will be described below with reference to the accompanying drawings. It should be understood that the preferred embodiments described herein are for illustration and explanation only and are not intended to limit the present invention.

[0050] The terminology used in the embodiments of this application is for the purpose of describing particular embodiments only and is not intended to limit the embodiments of this application. The singular forms “a,” “the,” and “the” used in the embodiments of this application are also intended to include the plural forms unless the context clearly indicates otherwise. It should also be understood that the term “and / or” as used herein refers to and includes any or all possible combinations of one or more associated listed items.

[0051] In the following description, when referring to the accompanying drawings, unless otherwise indicated, the same numbers in different drawings represent the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with this application. Rather, they are merely examples of apparatuses and methods consistent with some aspects of this application. In the description of this application, it should be understood that the terms "first," "second," "third," etc., are used only to distinguish similar objects and are not necessarily used to describe a specific order or sequence, nor should they be construed as indicating or implying relative importance. Those skilled in the art can understand the specific meaning of the above terms in this application according to the specific circumstances.

[0052] Example 1: As Figure 1 As shown, this embodiment of the invention provides an air quality sensor based on the Internet of Things, comprising:

[0053] The multi-source heterogeneous sensing fusion system is used to form a three-dimensional sensing layer through spatial grid deployment of a monitoring network composed of multiple sets of heterogeneous sensor nodes. Each heterogeneous sensor node collects the gaseous molecule concentration fluctuations in the physical space in real time, and obtains composite pollution characteristic factors through multi-source heterogeneous fusion processing. The composite pollution characteristic factors form a micro-environment quality evolution map. The gaseous molecule concentration fluctuations include air quality parameters such as PM2.5, CO2, TVOC, temperature, and humidity.

[0054] The spatiotemporal edge gateway system is used to deploy computing node clusters on the edge side of the air quality monitoring area. It adopts a hierarchical aggregation architecture to process the sensing data. Each edge node performs edge topology compression processing on the micro-environment map within its coverage area to obtain spatiotemporal feature slices. The spatiotemporal feature slices form a confined space pollution migration cloud strategy.

[0055] The dynamic response coordination hub system is used to process the cloud strategy for migrating pollution in enclosed spaces through a cross-domain strategy decoupling engine to obtain a set of device control vectors. The set of device control vectors forms a linkage control matrix for the purification devices. It pushes air quality exceeding alarms through a mobile app and links the device indicator lights to provide feedback on exceeding the standards.

[0056] The working principle and beneficial effects of the above technical solution are as follows: The multi-source heterogeneous sensing fusion system in this embodiment acquires the concentration fluctuations of gaseous molecules in physical space, and obtains composite pollution characteristic factors through multi-source heterogeneous fusion processing. These composite pollution characteristic factors form a microenvironment quality evolution map. The spatiotemporal domain edge gateway system obtains spatiotemporal feature slices from the microenvironment quality evolution map through edge topology compression processing. These spatiotemporal feature slices form a confined space pollution migration cloud strategy. The dynamic response collaborative hub system obtains a device control vector set from the confined space pollution migration cloud strategy through a cross-domain strategy decoupling engine. These device control vector sets form a purification device linkage control matrix. The above solution achieves a three-level technical closed loop in the field of IoT air quality monitoring: The multi-source heterogeneous sensing fusion system transforms discrete gaseous molecule concentration data into spatially correlated composite pollution characteristic factors, breaking through the limitations of traditional single-point measurement by sensors and constructing a quantifiable and analyzable microenvironment quality evolution map. The spatiotemporal domain edge system performs topology compression on high-dimensional map data to generate feature slices containing both temporal and spatial attributes, reducing the amount of original monitoring data while still fully characterizing the migration patterns of pollutants in confined spaces. The dynamic response coordination hub system decouples abstract migration strategies into executable device control vectors. The resulting control matrix enables coordinated responses of the purification device group, transforming traditionally independently operating purification devices into a spatially strategic interconnected network.

[0057] In summary, this embodiment transforms the pollution dynamics in the physical space into operable equipment control parameters through a three-stage data processing chain of sensing, calculation, and control, establishing a complete IoT closed-loop system from environmental monitoring to purification execution.

[0058] Example 2: Figure 2 As shown, based on Embodiment 1, the multi-source heterogeneous sensing fusion system provided in this embodiment of the invention includes...

[0059] The gaseous molecule concentration fluctuation acquisition subsystem is used to capture discrete gaseous oscillation parameters in physical space through a distributed resonant sensor array and generate a raw oscillation parameter stream. The raw oscillation parameter stream is decoupled by time-domain oscillation frequency and recalibrated by spatial gridding to eliminate the spatiotemporal reference offset between multi-source sensor units and output a synchronized oscillation tensor.

[0060] The composite pollution feature factor generation subsystem is used to perform multi-order differential manifold expansion on the synchronized oscillating tensor and extract gradient mutation features of each gaseous parameter; the gradient mutation features are fused by rotation-invariant convolution kernels to generate cross-modal pollution entropy values ​​with pollution correlation; the cross-modal pollution entropy values ​​are input into a chaotic topological clusterer to output composite pollution feature factors characterizing the intensity of composite pollution.

[0061] The microenvironment quality evolution map construction subsystem is used to obtain the pollution propagation potential energy between nodes by using composite pollution characteristic factors as nodes and an adaptive field strength program; a three-dimensional spatiotemporal connection topology network is constructed based on the pollution propagation potential energy, and a real-time pollution intensity scalar field is superimposed on the three-dimensional spatiotemporal connection topology network to finally form a microenvironment quality evolution map with pollution diffusion paths.

[0062] The working principle and beneficial effects of the above technical solution are as follows: The gaseous molecule concentration fluctuation acquisition subsystem of this embodiment is used to capture discrete gaseous oscillation parameters in physical space through a distributed resonant sensing array to generate a raw oscillation parameter stream; the raw oscillation parameter stream is decoupled by time-domain oscillation frequency and recalibrated by spatial gridding to eliminate the spatiotemporal reference offset between multi-source sensing units and output a synchronized oscillation tensor; the composite pollution feature factor generation subsystem is used to expand the synchronized oscillation tensor into a multi-order differential manifold and extract the gradient mutation features of each gaseous parameter; the gradient mutation features are fused by rotation-invariant convolution kernels to generate a cross-modal pollution entropy value with pollution correlation; the cross-modal pollution entropy value is input into a chaotic topological clusterer to output a composite pollution feature factor characterizing the intensity of composite pollution; the microenvironmental quality evolution map construction subsystem is used to obtain the pollution propagation potential energy between nodes through an adaptive field strength program, with the composite pollution feature factor as the node; a three-dimensional spatiotemporal connection topology network is constructed based on the pollution propagation potential energy, and a real-time pollution intensity scalar field is superimposed on the three-dimensional spatiotemporal connection topology network to finally form a microenvironmental quality evolution map with pollution diffusion paths. The above scheme uses discrete gaseous oscillation parameters captured by a distributed resonant sensing array to form a highly consistent oscillation tensor through spatiotemporal benchmark synchronization. Based on the cross-modal pollution entropy values ​​extracted using differential manifold unrolling and rotation-invariant convolution kernels, composite pollution feature factors are generated through chaotic topological clustering, achieving multi-parameter pollution correlation modeling. Finally, a three-dimensional spatiotemporal connected topology network constructed by an adaptive field strength program is superimposed with a real-time scalar field to form a micro-environmental quality evolution map with dynamic tracking capabilities for pollution diffusion paths. This achieves a complete process analysis from raw physical signal acquisition to visualization of pollution propagation trends. Its core value lies in establishing a multi-scale environmental quality monitoring framework of "physical signal-feature factor-spatiotemporal evolution." Through heterogeneous data fusion and topology modeling techniques, it overcomes the limitations of single-point discrete measurements in traditional environmental monitoring, providing a spatiotemporally continuous analytical basis for pollution source tracing and diffusion prediction in complex micro-environments.

[0063] Example 3: As Figure 3 As shown, based on Embodiment 1, the spatiotemporal domain edge gateway system provided in this embodiment of the invention includes:

[0064] The spatiotemporal dimension decomposition subsystem is used to extract the pollution propagation potential energy field and real-time pollution intensity scalar field from the microenvironment quality evolution map through four-dimensional tensor slicing, and generate a dual-channel spatiotemporal data stream.

[0065] The edge topology compression processing subsystem is used for processing dual-channel spatiotemporal data streams through an entropy-constrained collapse engine; it performs adaptive eddy detection on the pollution propagation potential energy field to capture the pollution migration direction vector; it performs spatial pyramid downsampling on the real-time pollution intensity scalar field to retain intensity gradient abrupt change points; and it generates high-information-density spatiotemporal feature slices by fusing the pollution migration direction vector and the intensity gradient abrupt change points through tensor cross product.

[0066] The edge topology compression processing subsystem is responsible for compressing the massive dual-channel data stream into high-information-density feature slices. The entropy-constrained collapse engine is a lossy compression framework whose goal is to maximize the retention of information useful for subsequent tasks within a given information entropy or complexity upper limit maximum allowable entropy value. The maximum allowable entropy value, in bits, is a hyperparameter determined by the computational and storage capabilities of the edge nodes. The smaller the maximum allowable entropy value, the higher the compression ratio, but the greater the potential distortion. The Lagrange multiplier λ is used in rate-distortion optimization to balance the compression ratio or bit rate R and the distortion degree or D. The engine automatically searches for the quantization scheme that minimizes J = D + λR, where J is the cost function. The engine quantizes and encodes the output results of subsequent vortex detection and downsampling operations to ensure that the data entropy of the final generated spatiotemporal feature slice does not exceed the maximum allowable entropy value.

[0067] The pollution migration cloud strategy construction subsystem is used to calculate the probability distribution of pollution sources through nonlinear diffusion inversion using spatiotemporal feature slices as input units; generate equipotential surfaces of pollution concentration and tangent bundles of migration paths based on the probability distribution of pollution sources; and encode the equipotential surfaces and tangent bundles of migration paths into a dynamic strategy topology network to form a closed-space pollution migration cloud strategy that can describe the law of pollution diffusion.

[0068] The pollution migration cloud strategy construction subsystem uses compressed feature slices to inversely infer pollution sources and predict their diffusion patterns; nonlinear diffusion inversion is used to calculate the probability distribution of pollution sources.

[0069] The forward model uses the convection-diffusion equation as the physical model to describe the evolution of pollution:

[0070]

[0071] S(x,y,z) is the pollution source term to be inverted;

[0072] The inversion problem is a typical inverse problem, given the concentration field C at a certain time t0. obs (From the information contained in the feature slice), we inversely find the source term S. Since the problem is ill-posed, regularization is required.

[0073] Objective function:

[0074]

[0075] F(S) represents the forward model operator, indicating the process of simulating the generation of the concentration field from source S; R(S) represents the regularization term, such as the L1 norm promoting sparse sources, the Total Variation norm promoting spatial continuity of the source, and λ ′ It is the regularization intensity parameter; by solving this optimization problem, the obtained S is the pollution source probability distribution. The higher the probability, the greater the possibility that the location is a real source.

[0076] The working principle and beneficial effects of the above technical solution are as follows: The spatiotemporal dimension decomposition subsystem of this embodiment is used to extract the pollution propagation potential energy field and the real-time pollution intensity scalar field from the microenvironment quality evolution map through four-dimensional tensor slicing, generating a dual-channel spatiotemporal data stream; the edge topology compression processing subsystem is used to process the dual-channel spatiotemporal data stream through an entropy constraint collapse engine; adaptive eddy current detection is performed on the pollution propagation potential energy field to capture the pollution migration direction vector; the real-time pollution intensity scalar field is downsampled by spatial pyramid to retain intensity gradient mutation points; high information density spatiotemporal feature slices are generated by tensor cross product fusion of the pollution migration direction vector and intensity gradient mutation points; the pollution migration cloud strategy construction subsystem is used to calculate the pollution source probability distribution through nonlinear diffusion inversion using the spatiotemporal feature slices as input units; pollution concentration equipotential surfaces and migration path tangent bundles are generated according to the pollution source probability distribution; the equipotential surfaces and migration path tangent bundles are encoded into a dynamic strategy topology network to form a closed-space pollution migration cloud strategy that can describe the pollution diffusion law. The above scheme achieves dynamic quantitative analysis and visual representation of the pollution diffusion patterns in confined spaces through multi-module collaborative processing. Specifically, it manifests in three progressive levels: The spatiotemporal dimension decomposition subsystem decouples the original environmental monitoring data into a pollution propagation potential energy field (vector field) and a real-time pollution intensity field (scalar field), forming a physically meaningful dual-channel spatiotemporal data stream, providing structured input for subsequent analysis. The edge topology compression subsystem preserves the pollution migration direction characteristics through eddy current detection, extracts intensity gradient abrupt change points using the spatial pyramid algorithm, and finally generates spatiotemporal feature slices through tensor operations. This process achieves data dimensionality reduction and feature enhancement. The pollution migration cloud strategy construction subsystem transforms discrete feature slices into a continuous diffusion model: it calculates the pollution source probability distribution through nonlinear inversion, generates equipotential surfaces describing the concentration gradient distribution, and quantifies the diffusion direction and rate using migration path tangent bundles; the final output dynamic strategy topology network can completely characterize the three-dimensional diffusion dynamics of pollutants in confined spaces.

[0077] In summary, this embodiment achieves end-to-end transformation from raw monitoring data to a computable diffusion model, providing quantitative evidence for pollutant source tracing, diffusion prediction, and prevention and control decisions. Error propagation across all modules is controlled within confidence intervals, ensuring the physical consistency of the output strategy.

[0078] Example 4: Based on Example 3, the edge topology compression processing subsystem provided in this embodiment of the invention includes:

[0079] The pollution propagation potential energy field processing component is used to scan the potential energy gradient distribution of the pollution propagation potential energy field through a variable-scale differential operator, extract the curl extrema of the potential energy change, and generate a pollution migration direction vector set containing scalars of direction angle and vortex intensity. The pollution migration direction vector set is processed by a vector field sparsification engine to remove noise vectors with vortex intensity below the dynamic threshold, and to perform curvature consistency clustering on adjacent vectors. The output is a topology-simplified pollution migration main direction vector bundle.

[0080] The core of the variable-scale differential operator's expression is the composite operation of the curl operator and the Gaussian smooth convolution, expressed as:

[0081]

[0082] Among them, Ω σ The variable-scale differential operator itself is a family of operators whose specific behavior is determined by the scale parameter σ; inputting a vector field into the operator yields another vector field as the output. The curl operator is a differential operator in vector calculus used to measure the rotational tendency and intensity of a vector field. It takes a vector field as input and outputs a new vector field representing its local rotation. The Gaussian kernel function is a smoothing function with the scale parameter σ as the standard; * indicates a convolution operation, which means smoothing the input vector field using the Gaussian kernel function; a scale is selected, and the potential energy field is smoothed by convolution using the Gaussian kernel at that scale to obtain the smoothed field; the curl of the smoothed field is calculated to obtain the vorticity field; the local extrema of the vorticity modulus are found in the output vorticity field, which are the candidate points of the pollution migration direction vector to be captured by the algorithm; a series of different scale parameter values ​​are traversed to analyze the potential energy field at different scales, and finally a robust set of pollution migration direction vectors is generated through a multi-scale fusion strategy;

[0083] The pollution intensity scalar field processing component is used to construct a second-order differential curvature map of the intensity field using a real-time pollution intensity scalar field, identify zero-crossing points in the curvature map as gradient abrupt change critical points, and establish a non-uniform sampling grid centered on the critical points. Through adaptive pyramid downsampling constrained by the critical points, intensity mean aggregation is performed in non-abrupt regions, and the original intensity values ​​are preserved with the gradient abrupt change critical points as the core, generating a set of pollution intensity gradient abrupt change points with weight labels.

[0084] The dual-channel data fusion component is used to convert the main direction vector bundle of pollution migration into a direction tensor field, map the set of abrupt change points of pollution intensity gradient into an intensity weight tensor, perform positive cross-multiplication of the two tensors, and output spatiotemporal feature slices that fuse the pollution migration direction and intensity abrupt change features.

[0085] The working principle and beneficial effects of the above technical solution are as follows: The pollution propagation potential energy field processing component in this embodiment is used to scan the potential energy gradient distribution of the pollution propagation potential energy field through a variable-scale differential operator, extract the curl extrema points of the potential energy change, and generate a pollution migration direction vector set, including direction angle and vortex intensity scalars; the pollution migration direction vector set is processed by a vector field sparsification engine to remove noise vectors with vortex intensity below the dynamic threshold, and curvature consistency clustering is performed on adjacent vectors; a topology-simplified pollution migration main direction vector bundle is output; the pollution intensity scalar field processing component is used to construct a binary index of the intensity field using a real-time pollution intensity scalar field. The system employs a differential curvature map, identifying zero-crossing points as gradient abrupt change critical points, and establishing a non-uniform sampling grid centered on these critical points. Through adaptive pyramid downsampling constrained by these critical points, intensity mean aggregation is performed in non-abrupt regions, preserving the original intensity values ​​around the gradient abrupt change critical points to generate a weighted set of pollution intensity gradient abrupt change points. A dual-channel data fusion component converts the main pollution migration direction vector bundle into a direction tensor field, maps the pollution intensity gradient abrupt change point set into an intensity weight tensor, performs a positive cross-multiplication operation between the two tensors, and outputs a spatiotemporal feature slice fusing pollution migration direction and intensity abrupt change features. The edge topology compression processing subsystem of the above scheme achieves efficient extraction and fusion of pollution diffusion features through modular collaborative work: the pollution propagation potential energy field processing component transforms the continuous potential energy field into a discrete set of migration direction vectors through differential operator scanning and curl analysis, and then obtains vector bundles reflecting the main diffusion paths through curvature clustering; this process completes the transformation from a continuous field to a discrete topological structure. The pollution intensity scalar field processing component utilizes second-order differential curvature detection and adaptive sampling to achieve data dimensionality reduction while preserving intensity abrupt change characteristics, generating a set of abrupt change points with weighted labels; this process maintains the topological representation of key intensity abrupt change information. The dual-channel data fusion component couples directional and intensity features through tensor operations, generating spatiotemporal feature slices that simultaneously contain pollutant migration direction information and intensity abrupt change information.

[0086] In summary, this embodiment achieves the following through cascaded processing: topological feature extraction of the original continuous field, screening and clustering of key feature points, tensor fusion of multimodal features, and finally outputs a compressed representation with both directional and intensity abrupt change features, providing a high-information-density feature representation for subsequent pollution diffusion analysis.

[0087] Example 5: Based on Example 4, the dual-channel data fusion component provided in this embodiment of the invention includes:

[0088] The Directional Vector Beam to Tensor Field Sub-component is used to convert the main directional vector beam of pollution migration into a unit directional basis by using the curvature consistency clustering center of the vector beam as the reference point and through tangent space mapping, thereby generating a set of directional basis beams.

[0089] The tensor field generation sub-component is used to convert the vortex intensity scalar of each directional basis of the directional basis set into directional weight coefficients, construct a three-dimensional directional tensor with the directional basis as the axis and the weight as the modulus, and output a pollution migration directional tensor field with spatial distribution properties.

[0090] The mutation point set to weighted tensor quantum component is used to establish a radial decay function centered on the critical point of the pollution intensity gradient mutation point set, calculate the intensity amplitude of the decay function based on the weight label of the point set, and generate a mutation kernel function distribution cloud; the vertices of the non-uniform sampling grid are used as tensor control points, and bicubic spline interpolation of the kernel function value is performed between the control points to output a pollution intensity weighted tensor field covering the entire space.

[0091] The working principle and beneficial effects of the above technical solution are as follows: The direction vector bundle to tensor field sub-component of this embodiment is used to convert the main direction vector bundle of pollution migration into a unit direction basis by taking the curvature consistency clustering center of the vector bundle as the reference point and converting each vector direction angle into a unit direction basis through tangent space mapping, thereby generating a set of direction basis; The tensor field generation sub-component is used to convert the vortex intensity scalar of each direction basis in the set of direction basis into direction weight coefficients, construct a three-dimensional direction tensor with the direction basis as the axis and the weight as the modulus, and output a pollution migration direction tensor field with spatial distribution attributes; The mutation point set to weight tensor quantum component is used to establish a radial decay function with the critical point as the center of the pollution intensity gradient mutation point set, calculate the intensity amplitude of the decay function according to the weight label of the point set, and generate a mutation kernel function distribution cloud; The vertices of the non-uniform sampling grid are used as tensor control points, and bicubic spline interpolation of the kernel function value is performed between the control points to output a pollution intensity weight tensor field covering the entire space. The dual-channel data fusion component of the above scheme achieves tensor-based representation and spatial fusion of pollution diffusion direction and intensity features through modular processing: the direction vector bundle to tensor field sub-component converts discrete pollution migration direction vector bundles into continuous direction tensor fields. Through curvature-consistent clustering and tangent space mapping, it ensures the geometric continuity of direction features while retaining vortex intensity as a weighting coefficient, enabling the tensor field to reflect the dominant trend of pollution diffusion. The mutation point set to weighted tensor quantum component converts discrete pollution intensity gradient mutation point sets into continuous spatial weighted tensor fields. Through radial decay functions and bicubic spline interpolation, it achieves a smooth transition while maintaining mutation features, ensuring the preservation of the topological structure of intensity changes. Tensor operation fusion performs a positive cross-multiplication operation between the direction tensor field and the intensity weighted tensor field, generating a spatiotemporal feature slice that fuses the pollution diffusion direction and intensity mutation features. This ensures that the final output can characterize both the migration path of pollutants and reflect the intensity change trend in key areas.

[0092] In summary, this embodiment realizes the conversion of discrete features into a continuous tensor field, and completes the coupling of multimodal information through tensor operations, providing a spatial feature expression that combines directionality and intensity variation for pollution diffusion analysis.

[0093] Example 6: Based on Example 5, the mutation point set to weighted tensor quantum component provided in this embodiment of the invention includes:

[0094] The function distribution cloud generation module is used to input the set of pollution intensity gradient abrupt change points into the radial decay function constructor, extract the spatial coordinates and weight label values ​​of each critical point, establish an exponential decay function with the critical point as the center, and dynamically adjust the function decay coefficient by the weight label value to generate a decay function unit with amplitude modulation.

[0095] The kernel function cloud synthesis module is used to superimpose the function units of all decay function units according to spatial coordinates, eliminate the interference effect in the overlapping area by amplitude normalization, and output a continuously distributed abrupt kernel function distribution cloud; the non-uniform sampling grid vertex input tensor control point binder reads the spatial coordinates of the grid vertex as the reference position of the control point, injects the kernel function amplitude at the corresponding position at the control point, and generates a tensor control point set with amplitude constraints;

[0096] The global tensor interpolation module is used to construct a three-dimensional parametric surface with control points as nodes, perform continuous interpolation of kernel function amplitude along the surface normal, ensure smooth gradient of the interpolated surface through curvature constraints, and output a pollution intensity weight tensor field covering the entire domain.

[0097] The working principle and beneficial effects of the above technical solution are as follows: The function distribution cloud generation module of this embodiment inputs the pollution intensity gradient mutation point set into the radial decay function constructor, extracts the spatial coordinates and weight label values ​​of each critical point, establishes an exponential decay function with the critical point as the center, and dynamically adjusts the function decay coefficient by the weight label value to generate a decay function unit with amplitude modulation; the kernel function cloud synthesis module superimposes each function unit of all decay function units according to spatial coordinates, eliminates the interference effect of overlapping areas by amplitude normalization, and outputs a continuously distributed mutation kernel function distribution cloud; the non-uniform sampling grid vertices are input into the tensor control point binder, the spatial coordinates of the grid vertices are read as the reference positions of the control points, and the kernel function amplitude of the corresponding position is injected at the control points to generate a tensor control point set with amplitude constraints; the global tensor interpolation module constructs a three-dimensional parametric surface with the control points as nodes, performs continuous interpolation of the kernel function amplitude along the surface normal, ensures the smoothness of the interpolation surface gradient by curvature constraints, and outputs a pollution intensity weight tensor field with global coverage. The above scheme, through mathematical modeling and spatial interpolation techniques, transforms a discrete set of pollution intensity abrupt change points into a continuous global weighted tensor field. Its main functions are: converting discrete abrupt change points into function units with amplitude modulation using a radial decay function constructor, each unit containing spatial coordinates, weighted labels, and a dynamic decay coefficient; and achieving a mapping from discrete observation data to a continuous function space. When synthesizing kernel function clouds using the superposition principle, L... 2 Norm normalization eliminates amplitude interference in overlapping regions, ensuring that the synthesized kernel function field satisfies physical constraints and avoiding non-physical phenomena such as amplitude overflow. In the parametric surface construction stage, bicubic B-spline basis functions are used for interpolation, and gradient smoothing is achieved by minimizing the curvature energy term. The final output tensor field satisfies the characteristics of elliptic partial differential equations, providing a weight distribution field conforming to differential geometry for subsequent physical simulations.

[0098] In summary, this embodiment constructs a complete mathematical transformation chain from discrete observation data, continuous function representation, to differentiable tensor fields, and its output can be directly used as the parameter field of non-uniform media in finite element analysis or computational fluid dynamics.

[0099] Example 7: Figure 4 As shown, based on Embodiment 1, the dynamic response collaborative hub system provided in this embodiment of the invention includes:

[0100] The migration cloud strategy decoupling subsystem is used to extract the pollution concentration equipotential surface features in the pollution migration cloud strategy, generate the pollution intensity gradient tensor, separate the curvature parameters of the tangent bundles of the migration path, generate the path direction admittance matrix, and output the decoupled strategy feature primitives; match the pollution intensity gradient tensor with the power response curve of the purification equipment, map the path direction admittance matrix to the equipment spatial orientation topology; and generate a dimensionally orthogonalized set of equipment control vectors.

[0101] The linkage control matrix construction subsystem is used to construct the equipment operation unitary space based on the equipment control vector, and to allocate the weight of the equipment control quantity in the unitary space according to the direction of the pollution intensity gradient; to establish the phase coordination relationship of the equipment control quantity according to the equipment spatial orientation topology, and to encode the weight allocation and phase coordination into a multi-dimensional control plane, and output the linkage control matrix of the purification device that can drive multiple terminals.

[0102] A dual-channel feedback triggering subsystem is used to synchronously activate the feedback arbitrator in conjunction with the control matrix;

[0103] In the mobile terminal channel, when the pollution intensity weight in the linkage control matrix exceeds the dynamic threshold: the spatial code of the exceeding area is extracted, and the spatial code and intensity weight are combined to form an over-threshold alarm vector, which is then pushed to the gateway as a gaseous parameter over-threshold alarm. After receiving the gaseous parameter over-threshold alarm, the gateway will link the playback, lighting, and ventilation systems according to a preset program, and the devices will display the content of the gaseous parameter over-threshold alarm according to a preset program. At the same time, the gateway will push the alarm information to the APP through the alarm information push platform. In the physical indication channel, based on the phase coordination relationship of the linkage control matrix, the key pollution nodes in the device orientation topology are analyzed, and the spectral warning pulse sequence of the corresponding device is triggered.

[0104] Among them, the dual-channel feedback triggering subsystem displays the alarm content for gaseous parameters exceeding the threshold according to a preset program, specifically including:

[0105] The device linkage alarm execution component is used to drive the spatial mapping of the device by the gaseous parameter over-threshold alarm vector. The over-threshold alarm vector output by the mobile terminal channel is input to the edge node or gateway. The edge node parses the spatial code in the vector and performs coordinate coupling with the device spatial orientation topology to generate a three-dimensional pollution focus positioning matrix.

[0106] The equipment control quantity phase reconstruction component is used to construct phase coordination relationship based on linkage control matrix, reconstruct equipment control parameters in positioning matrix, extract wavefront propagation vector of acoustic unit in equipment spatial orientation topology, analyze radiation angle distribution function of optical unit, obtain vortex generation curvature of gas flow field modulator, and output cross-medium field effect parameter set.

[0107] The multimodal alarm field generation component is used by edge execution nodes to convert field effect parameter sets into device commands; it generates a standing wave interference field pointing towards the pollution focus according to the wavefront propagation vector, forming an audible alarm path; it emits a pulsed spectral scanning beam according to the radiation angle distribution function to form an optical coordinate marker at the pollution focus; it generates a directional cleaning flow layer based on the curvature of the vortex generation, which advances along the migration path tangential beam to construct a spatially synchronized alarm field;

[0108] The information flow and physical field fusion component is used to convert the intensity weight in the over-threshold alarm vector into the acoustic field sound pressure gradient, which is dynamically modulated according to the weight value. The optical pulse frequency is positively correlated with the weight value, and the laminar flow velocity is graded according to the weight value to form a weight-field strength coupled response.

[0109] The alarm information digital synchronization component is used by edge execution nodes to decompose the over-threshold alarm vector into a spatial event fingerprint, including the pollution focus coordinates and intensity weights; field effect parameter summary, including standing wave mode, spectral encoding, and spherical curvature; push to the cloud strategy hub; after receiving it, the user terminal APP restores it into a dynamic model of the acoustic navigation path in three-dimensional space, the spatiotemporal evolution trajectory of the pollution focus spectral identifier, and a heat map of the clean spherical motion vector;

[0110] The pollution focus localization matrix transforms abstract spatial coding into an operable geometric coordinate system for equipment; the field effect parameter set transforms the phase relationship of equipment control quantities into physical field generation rules; the intensity weight value permeates the sound pressure gradient / pulse frequency / laminar velocity parameters, realizing a mathematical isomorphic expression of alarm intensity and pollution degree; the spatial event fingerprint fully preserves the spatiotemporal characteristics of the physical alarm field; and the user terminal accurately reconstructs the three-dimensional alarm scene through the field effect parameter summary.

[0111] The working principle and beneficial effects of the above technical solution are as follows: The migration cloud strategy decoupling subsystem of this embodiment is used to extract the equipotential surface features of pollution concentration in the pollution migration cloud strategy, generate the pollution intensity gradient tensor, separate the curvature parameters of the tangent bundle of the migration path, generate the path direction admittance matrix, and output the decoupled strategy feature primitive set; the pollution intensity gradient tensor is matched with the power response curve of the purification equipment, and the path direction admittance matrix is ​​mapped to the equipment spatial orientation topology; a dimensionally orthogonalized set of equipment control vectors is generated; the linkage control matrix construction subsystem is used to construct the equipment operation unitary space based on the equipment control vectors, and allocate the weights of the equipment control quantities in the unitary space according to the pollution intensity gradient direction; the phase coordination relationship of the equipment control quantities is established according to the equipment spatial orientation topology, and the weight allocation and phase coordination are encoded into multi-dimensional control. The system comprises a control plane that outputs a linkage control matrix capable of driving multiple terminals of purification devices; a dual-channel feedback trigger subsystem for synchronously activating the feedback arbitrator of the linkage control matrix; a mobile terminal channel that, when the pollution intensity weight in the linkage control matrix exceeds a dynamic threshold: extracts the spatial code of the exceeding area, synthesizes the spatial code and intensity weight into an over-threshold alarm vector, and pushes a gaseous parameter over-threshold alarm to the gateway; upon receiving the gaseous parameter over-threshold alarm, the gateway, according to a preset program, links playback, lighting, and ventilation systems, etc., and the devices display the content of the gaseous parameter over-threshold alarm according to a preset program; simultaneously, the gateway pushes the alarm information to the APP through an alarm information push platform; and a physical indication channel that, based on the phase coordination relationship of the linkage control matrix, analyzes key pollution nodes in the device orientation topology and triggers a spectral warning pulse sequence for the corresponding device. The dynamic response coordination central system of the above scheme, through mathematical modeling and control strategy integration, constructs a closed-loop control system for pollution migration and purification device linkage. First, the pollution migration characteristics are mathematically analyzed, concentration equipotential surface features are extracted to generate a gradient tensor, and the curvature parameters of the migration path are separated to construct a directional admittance matrix; then, the pollution characteristics are matched with the device response characteristics to establish an orthogonalized control vector set. At the control strategy level, an equipment operating space is constructed based on the equipment control vector. Control weights are allocated according to the direction of the pollution gradient, and a phase coordination relationship is established in conjunction with the equipment orientation topology. This is ultimately encoded into a multi-dimensional control plane, forming a linkage control matrix that can drive multiple terminal devices. The feedback mechanism adopts a dual-channel design: the mobile terminal channel monitors the pollution intensity weight, generates an alarm vector and pushes a warning information when the threshold is exceeded; the physical indicator channel analyzes key pollution nodes in the equipment topology and triggers spectral warning signals from the corresponding devices.

[0112] In summary, this embodiment achieves fully automated control of the entire process from pollution feature analysis and control strategy generation to equipment linkage response, ensuring dynamic optimization and precise intervention in the pollution control process.

[0113] Example 8: Based on Example 7, the linkage control matrix construction subsystem provided in this embodiment of the invention includes:

[0114] The phase coordination relationship establishment component is used to determine the spatial orientation topology of the equipment, calculate the shortest connection path length between any two equipment, convert the path length into the action delay time base, and generate a spatiotemporal delay field between the equipment. The spatiotemporal delay field generates a diffused wavefront with the pollution source location as the wave source. Based on the delay field, the phase offset angle of the control action of each equipment is calculated, and the equipment control phase difference vector is output.

[0115] The control tensor synthesis component is used to map the pollution intensity gradient weights to the control magnitude matrix and convert the phase difference vector into a time-series coordination operator.

[0116] The multidimensional plane construction component is used to construct a complex control plane with the magnitude matrix as the real part and the timing operator as the imaginary part. The complex plane is extended along the device space dimension to form a three-dimensional control tensor. Orthogonal projection compression is performed on the tensor to generate a multidimensional control plane.

[0117] The working principle and beneficial effects of the above technical solution are as follows: The phase coordination relationship establishment component of this embodiment is used to calculate the shortest connected path length between any two devices by taking the spatial orientation topology of the equipment, converting the path length into the action delay time base, and generating a spatiotemporal delay field between the devices; the spatiotemporal delay field generates a diffused wavefront with the pollution source location as the wave source, calculates the phase offset angle of the control action of each device according to the delay field, and outputs the device control phase difference vector; the control quantity tensor synthesis component is used to map the pollution intensity gradient weight into the control quantity amplitude matrix, and convert the phase difference vector into a timing coordination operator; the multidimensional plane construction component is used to construct a complex control plane with the amplitude matrix as the real part and the timing operator as the imaginary part, expands the complex plane along the device spatial dimension to form a three-dimensional control tensor volume, and performs orthogonal projection compression on the tensor volume to generate a multidimensional control plane. The aforementioned scheme's linkage control matrix construction subsystem achieves spatiotemporal collaborative optimization of pollution control equipment through a modular architecture. Its core significance lies in the following: the phase coordination component, through the spatiotemporal delay field calculated by the equipment topology path, transforms spatial distance quantities into precise time delay parameters, enabling a dynamic mapping relationship between the pollution diffusion wavefront and equipment response; the output phase difference vector quantifies the timing differences of each device's actions, providing a time reference for distributed control. The control quantity tensor component achieves nonlinear transformation between pollution intensity and control quantity through gradient weight mapping, while simultaneously encoding the phase difference vector into a timing operator, constructing a joint expression space for control intensity and timing; this tensor representation preserves the multidimensional coupling characteristics of control parameters. The multidimensional plane component constructs a hybrid parameter space (amplitude + timing) through the complex domain, and through orthogonal projection compression of a three-dimensional tensor, achieves dimensionality reduction mapping of high-dimensional control parameters to an operable plane; this process significantly reduces the complexity of the decision space while preserving the core control dimensions.

[0118] In summary, this embodiment achieves a closed-loop transformation from physical space topology to control parameter space. Through three-level processing of spatiotemporal delay field modeling, tensor synthesis, and dimensionality reduction projection, the pollution diffusion dynamics characteristics are transformed into an executable multi-device collaborative control strategy. This maintains the physical interpretability of the control parameters while providing an efficient collaborative decision-making mechanism.

[0119] Example 9: Based on Example 8, the multi-dimensional planar construction component provided in this embodiment of the invention includes:

[0120] The complex control plane construction sub-component is used to take each element of the control magnitude matrix as the complex real part and the corresponding element of the timing operator as the complex imaginary part to generate a set of complex control units for equipment. The set of complex control units for equipment establishes a complex plane coordinate system based on the equipment spatial orientation topology, maps the complex units to the complex plane according to the equipment coordinates, and outputs a complex control plane with equipment spatial association.

[0121] The 3D control tensor volume generation sub-component is used to extract the device position coordinates in the device spatial orientation topology from the complex control plane, with the position coordinates as the third dimension reference axis.

[0122] Tensor volume construction sub-component is used to replicate complex planes along the third dimension to form an initial tensor volume. The phase offset of each plane is adjusted according to the inter-device spatiotemporal delay field. The tensor volume surface is smoothed by curvature continuity constraints to generate a three-dimensional control tensor volume that carries the spatial phase relationship.

[0123] The working principle and beneficial effects of the above technical solution are as follows: In this embodiment, the complex control plane construction sub-component takes each element of the control magnitude matrix as the complex real part and the corresponding element of the timing operator as the complex imaginary part to generate a set of complex control units for equipment. The set of complex control units for equipment establishes a complex plane coordinate system based on the spatial orientation topology of the equipment, maps the complex units to the complex plane according to the equipment coordinates, and outputs a complex control plane with equipment spatial correlation. The three-dimensional control tensor volume generation sub-component extracts the equipment position coordinates in the spatial orientation topology of the equipment from the complex control plane, and takes the position coordinates as the third dimension reference axis. The tensor volume construction sub-component copies the complex plane along the third dimension to form an initial tensor volume, adjusts the phase offset of each plane according to the inter-equipment spatiotemporal delay field, smooths the surface of the tensor volume through curvature continuity constraints, and generates a three-dimensional control tensor volume that carries the spatial phase relationship. The core function of the multi-plane construction component in the above scheme lies in extending two-dimensional control parameters into a three-dimensional decision body with spatial phase characteristics. Its technical significance can be divided into two parts: The complex control plane construction sub-component combines the amplitude matrix (real part) and the timing operator (imaginary part) into a complex unit set, establishing a joint expression of control strength and action timing in the complex plane coordinate system. This representation method preserves the spatial topological relationship between devices, allowing each control command to simultaneously carry amplitude and timing information. The three-dimensional control tensor volume generation sub-component introduces device position coordinates as a third dimension, extending the complex plane along spatial orientation into an initial tensor volume; this achieves the dimensionality-up mapping of control parameters from a two-dimensional plane to three-dimensional space, providing a structured carrier for subsequent phase adjustment. The tensor volume construction sub-component dynamically adjusts the phase offset of each complex plane through a spatiotemporal delay field and applies curvature continuity constraints, ensuring a smooth transition on the tensor volume surface; the finally generated three-dimensional control tensor volume fully carries the spatial distribution and phase coordination relationship between devices, forming a control strategy space with spatiotemporal consistency.

[0124] In summary, this embodiment integrates discrete device control parameters into a continuously differentiable high-dimensional decision body through three steps: complex plane construction, three-dimensional tensor generation, and phase optimization. This provides a mathematical model foundation for multi-device collaborative control that combines spatial correlation and temporal coordination.

[0125] Example 10: Based on Example 9, the tensor body construction sub-component provided in this embodiment of the invention includes:

[0126] The initial tensor volume construction module is used to read the device position coordinate sequence in the device spatial orientation topology, copy the complex plane along the third dimension with the coordinate sequence as the index, and generate a discrete layered initial tensor volume.

[0127] The phase offset calibration module is used to initialize the spatiotemporal delay field between the tensor and the device, extract the delay time base of each device in the delay field, convert the time base into a complex plane rotation angle, perform a unitary transformation of the corresponding rotation angle on the complex plane of each device layer, and output a phase-synchronized transition tensor.

[0128] The curvature continuity processing module is used to obtain the curvature tensor of the tensor surface in the device dimension, align the eigenvalues ​​with the pollution intensity gradient tensor, optimize the surface continuity through the minimum curvature variational method, and generate a smooth three-dimensional control tensor with spatial phase relationship.

[0129] The working principle and beneficial effects of the above technical solution are as follows: The initial tensor body construction module of this embodiment is used to read the equipment position coordinate sequence in the equipment spatial orientation topology, and copy the complex plane along the third dimension with the coordinate sequence as the index to generate a discrete stacked initial tensor body; the phase offset calibration module is used to extract the time base of each equipment delay field between the initial tensor body and the equipment, convert the time base into the complex plane rotation angle, and perform a unitary transformation of the corresponding rotation angle on each equipment layer complex plane; output the phase-synchronized transition tensor body; the curvature continuity processing module is used to obtain the curvature tensor of the tensor body surface in the equipment dimension, align the eigenvalues ​​with the pollution intensity gradient tensor, optimize the surface continuity through the minimum curvature variational method, and generate a smooth three-dimensional control tensor body that bears the spatial phase relationship. The tensor volume construction sub-component of the above scheme achieves spatial continuity and phase synchronization optimization of control parameters through phased processing. Its technical significance lies in the following: The initial tensor volume construction module, through indexed replication of the device coordinate sequence, discretizes and stacks two-dimensional complex planes into a three-dimensional layered structure. This process preserves the integrity of the original control parameters and establishes an operable framework for the parameters along the spatial dimension, providing a basic topology for subsequent phase calibration. The phase offset calibration module converts the time basis of the spatiotemporal delay field into a complex plane rotation angle, and achieves phase synchronization adjustment of each device layer through unitary transformation. This operation enables the discretely distributed device control planes to form a coherent phase gradient in the complex domain, and the output transition tensor volume initially possesses spatiotemporal consistency. The curvature continuity processing module aligns the contamination gradient with the surface curvature features, constraining the tensor volume smoothness using the principle of least variation. The final generated three-dimensional control tensor volume mathematically satisfies differential geometric continuity and physically maintains the coupling relationship between contamination diffusion dynamics and control phase, forming a differentiable and physically interpretable high-dimensional control strategy space.

[0130] In summary, this embodiment realizes the transformation from discrete stacked structures to continuous smooth manifolds. Through coordinate indexed replication, unitary transformation phase calibration, and variational optimization, the control tensor volume simultaneously satisfies the three core requirements of topological correctness of spatial distribution, temporal consistency of phase coordination, and differential continuity of surface evolution.

[0131] Example 11: As Figure 5 As shown, based on Examples 1-10, the remote control method for an air quality sensor based on the Internet of Things provided in this embodiment of the invention includes the following steps:

[0132] S100: The purification device linkage control matrix is ​​processed by heterogeneous instruction fusion technology to obtain a spatially optimized control vector set, which forms a human-machine interaction entropy value mechanism.

[0133] S200: The human-computer interaction entropy value model is processed by the entropy weight dynamic allocation strategy to obtain the spatial state perception vector, and the spatial state perception vector forms a cross-media alarm triggering channel.

[0134] S300: The cross-media alarm triggering channel is processed by the bidirectional response coupling engine to obtain the device status feature topology, and the device status feature topology forms a closed-loop control entropy correction command.

[0135] The working principle and beneficial effects of the above technical solution are as follows: In this embodiment, firstly, the linkage control matrix of the purification device is processed by heterogeneous instruction fusion technology to obtain a spatial optimization control vector set, which forms a human-machine interaction entropy value mechanism; secondly, the human-machine interaction entropy value model is processed by the entropy weight dynamic allocation strategy to obtain a spatial state perception vector, which forms a cross-media alarm triggering channel; finally, the cross-media alarm triggering channel is processed by the bidirectional response coupling engine to obtain the equipment state feature topology, which forms a closed-loop control entropy correction instruction. The above solution achieves intelligent closed-loop management of air quality monitoring and purification control; it optimizes and integrates the control matrix of the purification device through heterogeneous command fusion technology to generate a spatially optimized control vector set, enabling multi-device collaborative control to have spatial adaptability, while establishing a human-machine interaction entropy value mechanism to achieve the interpretability of the control process; it adopts an entropy weight dynamic allocation strategy to process human-machine interaction data, generating a perception vector that accurately reflects changes in spatial state, and achieves multi-dimensional information synchronous transmission through a cross-media alarm triggering channel to ensure reliable transmission of alarm information; it constructs a device state characteristic topology network based on a bidirectional response coupling engine to form a closed-loop control system with feedback adjustment capabilities, and achieves dynamic optimization of the control strategy through entropy correction commands;

[0136] In summary, this embodiment realizes intelligent processing across the entire chain from environmental perception to equipment control, improving system response speed while ensuring control accuracy, enabling the air quality management system to have self-learning and self-adaptive capabilities, and forming a complete monitoring-analysis-control-optimization closed loop.

[0137] Obviously, those skilled in the art can make various modifications and variations to this invention without departing from its spirit and scope. Therefore, if these modifications and variations fall within the scope of equivalents of this invention, this invention is also intended to include these modifications and variations.

Claims

1. An air quality sensor based on the Internet of Things, characterized in that, Include: The multi-source heterogeneous sensing fusion system is used to form a three-dimensional sensing layer through spatial grid deployment of a monitoring network composed of multiple heterogeneous sensor nodes. Each heterogeneous sensor node collects the gaseous molecule concentration fluctuations in the physical space in real time, and obtains composite pollution characteristic factors through multi-source heterogeneous fusion processing. The composite pollution characteristic factors form a microenvironment quality evolution map. The spatiotemporal edge gateway system is used to deploy computing node clusters on the edge side of the air quality monitoring area. It adopts a hierarchical aggregation architecture to process the sensing data. Each edge node performs edge topology compression processing on the micro-environment map within its coverage area to obtain spatiotemporal feature slices. The spatiotemporal feature slices form a confined space pollution migration cloud strategy. The dynamic response coordination hub system is used to process the cloud strategy for migrating pollution in a confined space through a cross-domain strategy decoupling engine to obtain a set of equipment control vectors, which in turn form a linkage control matrix for the purification device. The spatiotemporal domain edge gateway system includes: The spatiotemporal dimension decomposition subsystem is used to extract the pollution propagation potential energy field and real-time pollution intensity scalar field from the microenvironment quality evolution map through four-dimensional tensor slicing, and generate a dual-channel spatiotemporal data stream. The edge topology compression processing subsystem is used for processing dual-channel spatiotemporal data streams through an entropy-constrained collapse engine; it performs adaptive eddy detection on the pollution propagation potential energy field to capture the pollution migration direction vector; it performs spatial pyramid downsampling on the real-time pollution intensity scalar field to retain intensity gradient abrupt change points; and it generates high-information-density spatiotemporal feature slices by fusing the pollution migration direction vector and the intensity gradient abrupt change points through tensor cross product. The pollution migration cloud strategy construction subsystem is used to calculate the probability distribution of pollution sources through nonlinear diffusion inversion using spatiotemporal feature slices as input units; generate equipotential surfaces of pollution concentration and tangent bundles of migration paths based on the probability distribution of pollution sources; and encode the equipotential surfaces and tangent bundles of migration paths into a dynamic strategy topology network to form a closed-space pollution migration cloud strategy that can describe the law of pollution diffusion. The edge topology compression processing subsystem includes: The pollution propagation potential energy field processing component is used to scan the potential energy gradient distribution of the pollution propagation potential energy field through a variable-scale differential operator, extract the curl extrema of the potential energy change, and generate a pollution migration direction vector set containing scalars of direction angle and vortex intensity. The pollution migration direction vector set is processed by a vector field sparsification engine to remove noise vectors with vortex intensity below the dynamic threshold, and to perform curvature consistency clustering on adjacent vectors. The output is a topology-simplified pollution migration main direction vector bundle. The pollution intensity scalar field processing component is used to construct a second-order differential curvature map of the intensity field using a real-time pollution intensity scalar field, identify zero-crossing points in the curvature map as gradient abrupt change critical points, and establish a non-uniform sampling grid centered on the critical points. Through adaptive pyramid downsampling constrained by the critical points, intensity mean aggregation is performed in non-abrupt regions, and the original intensity values ​​are preserved with the gradient abrupt change critical points as the core, generating a set of pollution intensity gradient abrupt change points with weight labels. The dual-channel data fusion component is used to convert the main direction vector bundle of pollution migration into a direction tensor field, map the set of abrupt change points of pollution intensity gradient into an intensity weight tensor, perform positive cross-multiplication of the two tensors, and output spatiotemporal feature slices that fuse the pollution migration direction and intensity abrupt change features.

2. The air quality sensor based on the Internet of Things as described in claim 1, characterized in that, The dual-channel data fusion component includes: The Directional Vector Beam to Tensor Field Sub-component is used to convert the main directional vector beam of pollution migration into a unit directional basis by using the curvature consistency clustering center of the vector beam as the reference point and through tangent space mapping, thereby generating a set of directional basis beams. The tensor field generation sub-component is used to convert the vortex intensity scalar of each directional basis of the directional basis set into directional weight coefficients, construct a three-dimensional directional tensor with the directional basis as the axis and the weight as the modulus, and output a pollution migration directional tensor field with spatial distribution properties. The mutation point set to weighted tensor quantum component is used to establish a radial decay function centered on the critical point of the pollution intensity gradient mutation point set, calculate the intensity amplitude of the decay function based on the weight label of the point set, and generate a mutation kernel function distribution cloud; the vertices of the non-uniform sampling grid are used as tensor control points, and bicubic spline interpolation of the kernel function value is performed between the control points to output a pollution intensity weighted tensor field covering the entire space.

3. The air quality sensor based on the Internet of Things as described in claim 2, characterized in that, The mutation point set to weighted tensor quantum component includes: The function distribution cloud generation module is used to input the set of pollution intensity gradient abrupt change points into the radial decay function constructor, extract the spatial coordinates and weight label values ​​of each critical point, establish an exponential decay function with the critical point as the center, and dynamically adjust the function decay coefficient by the weight label value to generate a decay function unit with amplitude modulation. The kernel function cloud synthesis module is used to superimpose the function units of all decay function units according to spatial coordinates, eliminate the interference effect in the overlapping area by amplitude normalization, and output a continuously distributed abrupt kernel function distribution cloud; the non-uniform sampling grid vertex input tensor control point binder reads the spatial coordinates of the grid vertex as the reference position of the control point, injects the kernel function amplitude at the corresponding position at the control point, and generates a tensor control point set with amplitude constraints; The global tensor interpolation module is used to construct a three-dimensional parametric surface with control points as nodes, perform continuous interpolation of kernel function amplitude along the surface normal, ensure smooth gradient of the interpolated surface through curvature constraints, and output a pollution intensity weight tensor field covering the entire domain.

4. The air quality sensor based on the Internet of Things as described in claim 1, characterized in that, Dynamic response coordination central system, including: The migration cloud strategy decoupling subsystem is used to extract the pollution concentration equipotential surface features in the pollution migration cloud strategy, generate the pollution intensity gradient tensor, separate the curvature parameters of the tangent bundle of the migration path, generate the path direction admittance matrix, and output the decoupled strategy feature primitive set. The pollution intensity gradient tensor is matched with the power response curve of the purification equipment, and the path direction admittance matrix is ​​mapped to the spatial orientation topology of the equipment; a set of equipment control vectors with orthogonal dimensions is generated. The linkage control matrix construction subsystem is used to construct the equipment operation unitary space based on the equipment control vector, and to allocate the weight of the equipment control quantity in the unitary space according to the direction of the pollution intensity gradient. Based on the spatial orientation topology of the equipment, the phase coordination relationship of the equipment control quantities is established, and the weight allocation and phase coordination are encoded into a multi-dimensional control plane, outputting a purification device linkage control matrix that can drive multiple terminals. A dual-channel feedback triggering subsystem is used to synchronously activate the feedback arbitrator in conjunction with the control matrix.

5. The air quality sensor based on the Internet of Things as described in claim 4, characterized in that, The dual-channel feedback triggering subsystem includes: a mobile terminal channel; when the pollution intensity weight in the linkage control matrix exceeds the dynamic threshold: extracting the spatial code of the exceeding area, synthesizing the spatial code and intensity weight into an over-threshold alarm vector, and pushing a gaseous parameter over-threshold alarm to the gateway; after receiving the gaseous parameter over-threshold alarm, the gateway will link the playback, lighting, and ventilation system devices according to a preset program, and the devices will display the content of the gaseous parameter over-threshold alarm according to a preset program; at the same time, the gateway will push the alarm to the APP through the alarm information push platform. The physical indicator channel analyzes key pollution nodes in the equipment orientation topology based on the phase coordination relationship of the linkage control matrix, and triggers the spectral warning pulse sequence of the corresponding equipment.

6. The air quality sensor based on the Internet of Things as described in claim 5, characterized in that, The linkage control matrix construction subsystem includes: The phase coordination relationship establishment component is used to determine the spatial orientation topology of the equipment, calculate the shortest connection path length between any two equipment, convert the path length into the action delay time base, and generate a spatiotemporal delay field between the equipment. The spatiotemporal delay field generates a diffused wavefront with the pollution source location as the wave source. Based on the delay field, the phase offset angle of the control action of each equipment is calculated, and the equipment control phase difference vector is output. The control tensor synthesis component is used to map the pollution intensity gradient weights to the control magnitude matrix and convert the phase difference vector into a time-series coordination operator. The multidimensional plane construction component is used to construct a complex control plane with the magnitude matrix as the real part and the timing operator as the imaginary part. The complex plane is extended along the device space dimension to form a three-dimensional control tensor. Orthogonal projection compression is performed on the tensor to generate a multidimensional control plane.

7. The air quality sensor based on the Internet of Things as described in claim 1, characterized in that, Fluctuations in gaseous molecule concentration include air quality parameters such as PM2.5, CO2, TVOC, temperature, and humidity.

Images