A hot and humid multi-channel acquisition method, device and medium based on edge computing

Abstract

The application discloses a kind of based on edge computing's hot and humid multi-channel acquisition method, equipment and medium, it is related to edge computing technical field, including, to gait phase edge collection benchmark set is locked phase synchronous acquisition and phase alignment processing, generates locked phase hot and humid original data stream;To locked phase hot and humid original data stream is preprocessed in edge, generates regional preprocessed hot and humid data, and extracts local hot and humid change rule to carry out time series deduction and contrast verification, generates regional hot and humid prediction contrast set;According to regional hot and humid prediction contrast set, identify edge deviation and carry out fragment classification processing, generate edge residual feature package;According to edge residual feature package, fusion each regional time series information and reconstruct whole body hot and humid state, generate whole body hot and humid fusion result.The application improves the synchronism and accuracy of multi-channel hot and humid data acquisition under dynamic test condition, provides reliable data basis for high-precision evaluation clothing hot and humid comfort.

Description

Technical Field

[0001] This invention relates to the field of edge computing technology, and in particular to a method, device and medium for multi-channel heat and humidity acquisition based on edge computing. Background Technology

[0002] With the development of clothing ergonomics testing, thermal and moisture comfort evaluation, and warm body dummy control technology, multi-anatomical area sensing deployment, zoned heating control, local sweating simulation, and dynamic posture driving are gradually becoming integrated. Thermal and moisture multi-channel acquisition systems are beginning to evolve from single-point measurement to multi-area synchronous sensing, and from offline analysis to real-time computing and processing.

[0003] Existing thermal and humidity acquisition schemes mostly adopt centralized aggregation and unified processing methods. Under dynamic attitude switching and multi-channel concurrent acquisition conditions, they lack acquisition task orchestration, unified timing, phase synchronization, time series extrapolation and verification, and backhaul update control mechanisms for multiple edge nodes. This easily leads to problems such as cross-regional time series mismatch, local response lag, and insufficient representation of high-frequency change segments, which in turn affects the continuity of thermal and humidity state reconstruction and the pertinence of acquisition strategy adjustment. Therefore, it is urgent to build a new acquisition method around edge-side time series coordination and local thermal and humidity evolution modeling. Summary of the Invention

[0004] In view of the aforementioned existing problems, the present invention is proposed.

[0005] Therefore, this invention provides a thermal and humidity multi-channel acquisition method based on edge computing to solve the problems of timing mismatch in thermal and humidity multi-channel acquisition and insufficient real-time collaborative processing at the edge side under dynamic testing conditions.

[0006] To solve the above-mentioned technical problems, the present invention provides the following technical solution:

[0007] In a first aspect, the present invention provides a multi-channel thermal and humidity acquisition method based on edge computing, comprising: acquiring regional distribution data and motion state data, and performing edge node deployment and unified timing processing to generate a gait phase edge acquisition reference set; performing phase-locked synchronous acquisition and phase alignment processing on the gait phase edge acquisition reference set to generate a phase-locked thermal and humidity raw data stream; performing edge preprocessing on the phase-locked thermal and humidity raw data stream to generate regional preprocessed thermal and humidity data, and extracting local thermal and humidity change patterns for time-series extrapolation and comparison verification to generate a regional thermal and humidity prediction reference set; identifying edge deviations and performing segment-level processing based on the regional thermal and humidity prediction reference set to generate edge residual feature packets; fusing temporal information of each region and reconstructing the whole-body thermal and humidity state based on the edge residual feature packets to generate a whole-body thermal and humidity fusion result; and performing acquisition strategy correction and backhaul update processing on the whole-body thermal and humidity fusion result to generate an updated gait phase edge acquisition reference set.

[0008] As a preferred embodiment of the edge-computation-based multi-channel thermal and humidity acquisition method of the present invention, the specific steps for generating the gait phase edge acquisition reference set are as follows:

[0009] Information on the anatomical boundaries, zoned heating distribution, localized sweating distribution, and joint movement trajectories of the warm-body dummy were collected and organized according to a unified time reference to generate the original regional motion dataset.

[0010] Based on the original regional motion dataset, spatial correspondence and temporal association are performed on the anatomical region boundary information, zoned heating distribution information, local sweating distribution information and joint motion trajectory information to generate regional distribution data and motion state data.

[0011] Based on regional distribution data and motion status data, the jurisdiction of edge nodes is divided and access relationships are established. Unified timing is performed in combination with the order of attitude switching, and the edge node deployment timing result is generated.

[0012] Based on the timing results of edge node deployment, the corresponding sampling start time, sampling duration, phase marking rules and upload trigger conditions are configured for the jurisdiction of each edge node, generating a gait phase edge acquisition reference set.

[0013] As a preferred embodiment of the edge computing-based multi-channel thermal and humidity data acquisition method of the present invention, the specific steps for generating the phase-locked loop thermal and humidity raw data stream are as follows:

[0014] Based on the gait phase edge acquisition reference set, the sampling start time, sampling duration, phase marking rules and upload trigger conditions of the area under the jurisdiction of each edge node are sent to the corresponding edge node to generate a phase-locked acquisition control set;

[0015] According to the phase-locked acquisition control set, each edge node synchronously acquires temperature information, humidity information, zone heating status information, local sweating status information and joint motion status information within the corresponding phase window, and performs edge association to generate a regional phase-locked acquisition segment set.

[0016] Based on the regional phase-locked acquisition fragment set, the temperature information, humidity information, zone heating status information, local sweating status information and joint movement status information within the jurisdiction of each edge node are corrected for local time offset according to the phase marking rules, and then spliced ​​and organized according to a unified time order and phase sequence to generate the phase-locked thermal and humidity raw data stream.

[0017] As a preferred embodiment of the edge computing-based multi-channel thermal and humidity data acquisition method of the present invention, the specific steps for preprocessing the thermal and humidity data in the generation area are as follows:

[0018] By utilizing the jurisdiction of edge nodes and phase marking rules, the original phase-locked thermal and humidity data stream is segmented and correlated to generate a regional thermal and humidity segmented dataset;

[0019] Based on the continuity of change between adjacent phase windows, drift correction, time correction, anomaly identification and missing completion are performed on the regional thermal and humidity segmented dataset to generate a regional cleaning thermal and humidity dataset.

[0020] The regional cleaning thermal and humidity dataset is standardized in format and ordered to generate regional preprocessed thermal and humidity data.

[0021] As a preferred embodiment of the edge computing-based multi-channel thermal and humidity acquisition method of the present invention, the specific steps for generating the regional thermal and humidity prediction reference set are as follows:

[0022] The change trajectories of temperature, humidity, zone heating status, local sweating status and joint movement status within a continuous phase window are extracted from the preprocessed regional thermal and humidity data, and correlation analysis is performed to generate local thermal and humidity change patterns.

[0023] Based on the local heat and humidity variation patterns and the preprocessed heat and humidity data of the region, the temperature information, humidity information, zone heating status information, local sweating status information and joint movement status information in the next phase window are predicted to generate a set of regional heat and humidity prediction results.

[0024] The regional thermal and humidity prediction result set is matched and compared with the regional preprocessed thermal and humidity data to generate a regional thermal and humidity prediction comparison set.

[0025] As a preferred embodiment of the edge-computation-based multi-channel thermal and humidity acquisition method of the present invention, the specific steps for generating the edge residual feature packet are as follows:

[0026] Prediction difference information and temporal change information are extracted from the regional heat and humidity prediction control set, and then aggregated according to the jurisdiction of the edge nodes to generate a regional deviation analysis dataset.

[0027] Based on the regional deviation analysis dataset, we perform continuity analysis and deviation intensity analysis on the prediction difference information and temporal change information within the jurisdiction of each edge node, identify edge deviation situations and label corresponding segments, and generate a regional segment hierarchical dataset.

[0028] Based on the regional segment hierarchical dataset, the prediction difference information, temporal change information, phase labeling information and regional identification information of different hierarchical segments are aggregated and encapsulated to generate edge residual feature packages.

[0029] As a preferred embodiment of the edge computing-based multi-channel thermal and humidity acquisition method described in this invention, the specific steps for generating whole-body thermal and humidity fusion results are as follows:

[0030] Based on phase marker information and region identifier information, the edge residual feature packets are temporally aggregated to generate a regional temporal fusion dataset.

[0031] Cross-regional alignment and continuous stitching are performed on the regional temporal fusion dataset to generate a whole-body thermal and humidity reconstruction dataset;

[0032] The whole-body thermo-humidity reconstruction dataset is correlated and its state is reconstructed to generate the whole-body thermo-humidity fusion result.

[0033] As a preferred embodiment of the edge-computation-based multi-channel thermal and humidity acquisition method of the present invention, the specific steps for generating the updated gait phase edge acquisition reference set are as follows:

[0034] The thermal and moisture deviation distribution information, phase response difference information, and temporal change intensity information of different edge node jurisdiction areas are extracted from the whole-body thermal and moisture fusion results, and then collected according to the regional identification information and phase label information to generate a set of acquisition strategy correction information.

[0035] Based on the acquisition strategy, the information set is modified, and the sampling start time, phase marking rules, and upload triggering conditions of the jurisdiction of each edge node are adjusted to generate an edge acquisition update information set.

[0036] The edge acquisition update information set is sent back to the corresponding edge node jurisdiction area, and the gait phase edge acquisition reference set is updated to generate the updated gait phase edge acquisition reference set.

[0037] In a second aspect, the present invention provides a computer device including a memory and a processor, wherein the memory stores a computer program, wherein when the computer program is executed by the processor, it implements any step of the edge computing-based multi-channel thermal and humidity acquisition method described in the first aspect of the present invention.

[0038] Thirdly, the present invention provides a computer-readable storage medium having a computer program stored thereon, wherein: when the computer program is executed by a processor, it implements any step of the edge computing-based multi-channel thermal and humidity acquisition method described in the first aspect of the present invention.

[0039] The beneficial effects of this invention are as follows: by identifying edge deviations through a regional thermal and humidity prediction control set and performing segment hierarchical processing to generate edge residual feature packages, the continuous thermal and humidity change process can be transformed into a structured representation result with deviation intensity, temporal location and regional affiliation at the edge computing node side. This allows for the centralized expression of the coupling features between local temperature and humidity changes, zoned heating response, local sweating response and motion state perturbation, and forms a high information density data carrier for subsequent cross-regional temporal fusion, thereby improving the temporal coherence, regional identification and update responsiveness in the whole-body thermal and humidity state reconstruction process. Attached Figure Description

[0040] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the following description of the embodiments will be briefly introduced. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0041] Figure 1 This is a flowchart of a multi-channel thermal and humidity acquisition method based on edge computing.

[0042] Figure 2 This is a flowchart for generating the raw data stream of phase-locked loop thermo-humidity data.

[0043] Figure 3 A flowchart for generating edge residual feature packets.

[0044] Figure 4 The flowchart shows the generation of the updated gait phase edge acquisition reference set. Detailed Implementation

[0045] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings.

[0046] Many specific details are set forth in the following description in order to provide a full understanding of the invention. However, the invention may also be practiced in other ways different from those described herein, and those skilled in the art can make similar extensions without departing from the spirit of the invention. Therefore, the invention is not limited to the specific embodiments disclosed below.

[0047] Secondly, the term "one embodiment" or "embodiment" as used herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The phrase "in one embodiment" appearing in different places in this specification does not necessarily refer to the same embodiment, nor is it a single or selective embodiment that is mutually exclusive with other embodiments.

[0048] Reference Figures 1-4This is one embodiment of the present invention, which provides a multi-channel thermal and humidity acquisition method based on edge computing, including the following steps:

[0049] S1. Collect regional distribution data and motion state data, and perform edge node deployment and unified timing processing to generate a gait phase edge acquisition reference set.

[0050] S1.1 Collect anatomical region boundary information, zone heating distribution information, local sweating distribution information, and joint movement trajectory information of the warm-body dummy, and organize them according to a unified time reference to generate the original regional motion dataset.

[0051] Specifically, anatomical region boundary information is collected from the pre-defined anatomical partition marking positions on the surface of the warm-body dummy. The anatomical region boundary information is formed by reading the boundary coordinates of each anatomical partition. Partition heating distribution information is collected from the heating positions and connection positions of each partition of the warm-body dummy. The partition heating distribution information is formed by reading the correspondence between the heating positions of each partition and the anatomical partition. Local sweating distribution information is collected from the sweating positions and fluid supply positions on the surface of the warm-body dummy. The local sweating distribution information is formed by reading the correspondence between the sweating positions and the anatomical partition. Joint motion trajectory information is collected from the position and angle changes of each active joint of the warm-body dummy during continuous movement. The anatomical region boundary information, partition heating distribution information, local sweating distribution information, and joint motion trajectory information are time-stamped, sequentially aligned, and aggregated according to a unified time reference to generate the original region motion dataset.

[0052] It should also be noted that the preset anatomical zone marking positions refer to the marking points or marking areas that are pre-set according to the boundary division of each anatomical region of the warm body dummy and used to identify the spatial position of the corresponding anatomical zone.

[0053] A unified time reference refers to a unified reference time standard used when uniformly marking and aligning information on anatomical region boundaries, zonal heating distribution, local sweating distribution, and joint movement trajectories.

[0054] S1.2. Based on the original regional motion dataset, spatial correspondence and temporal correlation are performed on the anatomical region boundary information, zoned heating distribution information, local sweating distribution information, and joint motion trajectory information to generate regional distribution data and motion state data.

[0055] Specifically, anatomical region boundary information is extracted from the original regional motion dataset, and the spatial range of each anatomical region is delineated based on the anatomical region boundary information. The regional heating distribution information and local sweating distribution information are mapped to the corresponding anatomical region spatial range to form the corresponding thermal and moisture distribution positional relationship of each anatomical region. The change process of joint motion trajectory information at continuous time points is read according to a unified time reference, and the joint motion trajectory information is correlated and temporally associated with the corresponding anatomical region boundary information, regional heating distribution information, and local sweating distribution information. This makes the spatial positional relationship and motion change relationship of each anatomical region correspond to each other at different motion times, generating regional distribution data and motion state data.

[0056] It should also be noted that the spatial range of the anatomical region refers to the boundary position and boundary coordinates corresponding to the boundary information of the anatomical region, such as the range of the head region, chest region, back region, upper limb region, lower limb region, and perineal region.

[0057] S1.3. Based on regional distribution data and motion status data, divide the jurisdiction area of ​​the edge nodes and establish access relationships. Combine the order of attitude switching to perform unified timing and generate edge node deployment timing results.

[0058] Specifically, the boundary coordinates, zone heating mapping positions, and local sweating mapping positions of each anatomical region are read from the regional distribution data. Anatomical regions with adjacent boundaries and continuously covered by the same edge node are selected as the same candidate region. When the temperature and humidity acquisition positions of adjacent anatomical regions are continuously adjacent within the same posture switching cycle, and the zone heating distribution information and local sweating distribution information have a continuous coverage relationship, the corresponding anatomical region is divided into the edge node jurisdiction area. The edge node jurisdiction area is matched one by one with the temperature acquisition position, humidity acquisition position, zone heating distribution information, local sweating distribution information, and joint motion trajectory information within the corresponding anatomical region to establish the access relationship between the edge node jurisdiction area and the acquisition content. The posture switching time and movement sequence are extracted from the motion state data, and the acquisition time corresponding to each edge node jurisdiction area is uniformly synchronized and arranged in sequence according to the posture switching sequence, so that the edge node jurisdiction area, access relationship, and timing sequence correspond to each other, and the edge node deployment timing result is generated.

[0059] It should also be noted that the order of posture switching refers to the sequence of occurrence and the connection between different posture changes on the continuous time axis in the joint motion trajectory information.

[0060] S1.4. Based on the timing results of edge node deployment, configure the corresponding sampling start time, sampling duration, phase marking rules and upload trigger conditions for the jurisdiction of each edge node, and generate a gait phase edge acquisition reference set.

[0061] Specifically, the timing sequence and acquisition time information corresponding to the jurisdiction of each edge node are extracted from the timing results of the edge node deployment. The temporal and positional relationships of the jurisdiction of each edge node during continuous posture changes are read according to the timing sequence. The sampling start time and sampling duration are set for each jurisdiction of each edge node according to the temporal and positional relationships. The posture switching time is identified according to the joint motion trajectory information. The continuous sampling interval is divided into multiple consecutive phase intervals according to the order of posture switching. The corresponding posture stage is then marked for each phase interval, forming the phase division result corresponding to each jurisdiction of each edge node. The phase marking rules are configured for each jurisdiction of each edge node according to the phase division result. The upload trigger conditions are configured according to the order and temporal strength of the thermal and humidity changes in different phases of each jurisdiction of each edge node, so that the sampling start time, sampling duration, phase marking rules and upload trigger conditions correspond one-to-one in each jurisdiction of each edge node, generating a gait phase edge acquisition reference set.

[0062] It should also be noted that the sampling start time refers to the time point at which each edge node's jurisdiction begins to collect thermal and humidity information; the sampling duration refers to the length of time that each edge node's jurisdiction continuously collects thermal and humidity information within the corresponding phase; the phase marking rule refers to the rule for distinguishing and identifying different sampling stages in the continuous attitude change process; and the upload trigger condition refers to the judgment condition for starting data upload when the edge node's jurisdiction meets the timing status or thermal and humidity change requirements. Among these, the timing status or thermal and humidity change requirements refer to the situation where each edge node's jurisdiction reaches the preset phase switching time and the end time of the sampling stage, or when the temperature change amplitude, humidity change amplitude, and thermal and humidity change rate reach the preset upload standard.

[0063] S2. Perform phase-locked synchronous acquisition and phase alignment processing on the gait phase edge acquisition reference set to generate phase-locked thermal and humidity raw data stream.

[0064] S2.1 Based on the gait phase edge acquisition reference set, the sampling start time, sampling duration, phase marking rules, and upload trigger conditions of the area under the jurisdiction of each edge node are sent to the corresponding edge node to generate a phase-locked acquisition control set.

[0065] Specifically, the sampling start time, sampling duration, phase marking rules, and upload triggering conditions corresponding to the jurisdiction of each edge node are extracted from the gait phase edge acquisition reference set. Based on the correspondence between the jurisdiction of each edge node and the edge node itself, the sampling start time, sampling duration, phase marking rules, and upload triggering conditions are written into the acquisition control content of the corresponding edge node. Sampling time constraints are formed for each edge node's jurisdiction based on the sampling start time and sampling duration; phase recognition content is formed for each edge node's jurisdiction based on the phase marking rules; and upload decision content is formed for each edge node's jurisdiction based on the upload triggering conditions. This ensures that the sampling time constraints, phase recognition content, and upload decision content correspond to each other within the same edge node's jurisdiction, generating a phase-locked acquisition control set.

[0066] S2.2 According to the phase-locked acquisition control set, each edge node synchronously acquires temperature information, humidity information, zone heating status information, local sweating status information and joint motion status information within the corresponding phase window, and performs edge association to generate a regional phase-locked acquisition segment set.

[0067] Specifically, based on the sampling start time and sampling duration corresponding to the jurisdiction area of ​​each edge node in the phase-locked acquisition control set, temperature information, humidity information, zone heating status information, local sweating status information, and joint motion status information are collected in the corresponding phase window. Phase markers are added to the temperature information, humidity information, zone heating status information, local sweating status information, and joint motion status information according to the phase marking rules. Based on the upload trigger conditions, the temperature information, humidity information, zone heating status information, local sweating status information, and joint motion status information with the same time and position relationship in the same phase window are correspondingly organized and edge-associated, so that the temperature information, humidity information, zone heating status information, local sweating status information, and joint motion status information form a one-to-one corresponding phase acquisition content in the jurisdiction area of ​​each edge node, generating a regional phase-locked acquisition segment set.

[0068] S2.3. Based on the regional phase-locked acquisition fragment set, the temperature information, humidity information, zone heating status information, local sweating status information and joint movement status information within the jurisdiction of each edge node are corrected for local time offset according to the phase marking rules, and then spliced ​​and organized according to a unified time order and phase sequence to generate the phase-locked thermal and humidity raw data stream.

[0069] Specifically, the temperature, humidity, zone heating status, local sweating status, and joint motion status information corresponding to each edge node's jurisdiction within each phase window are extracted from the regional phase-locked acquisition segment set. The phase marker content and time position content corresponding to each phase window are also read. Based on the phase marker rules, local time offset corrections are applied to the temperature, humidity, zone heating status, local sweating status, and joint motion status information between adjacent phase windows within each edge node's jurisdiction, ensuring that these information maintain a time correspondence at the same phase position and at the intersection of adjacent phases. Following a unified time order and phase sequence, the temperature, humidity, zone heating status, local sweating status, and joint motion status information that have undergone local time offset correction are spliced ​​and continuously processed to generate a phase-locked thermal and humidity raw data stream.

[0070] S3. Perform edge preprocessing on the phase-locked thermal and humidity raw data stream to generate regional preprocessed thermal and humidity data, and extract local thermal and humidity variation patterns for time series extrapolation and comparison verification to generate a regional thermal and humidity prediction comparison set.

[0071] S3.1. Using the jurisdiction of edge nodes and phase marking rules, segment and associate the phase-locked thermal and humidity raw data stream to generate a regional thermal and humidity segmented dataset.

[0072] Specifically, temperature, humidity, zone heating status, local sweating status, and joint motion status information corresponding to the jurisdiction of each edge node are extracted from the phase-locked loop (PLL) thermal and humidity raw data stream. Based on the edge node jurisdiction division results, anatomical region correspondence, node identification information, and time stamp information, the information records corresponding to temperature, humidity, zone heating status, local sweating status, and joint motion status are classified into regions to form region classification results. The phase stamp content corresponding to each time position in the region classification results is parsed according to the phase stamp rules, and the temperature, humidity, zone heating status, local sweating status, and joint motion status information within the same phase interval in the jurisdiction of the same edge node are divided into corresponding phase data segments according to the phase stamp content. The temperature, humidity, zone heating status, local sweating status, and joint motion status information in each phase data segment are established in chronological order to generate a regional thermal and humidity segmented dataset.

[0073] S3.2 Based on the continuity of changes between adjacent phase windows, perform drift correction, time correction, anomaly identification, and missing data completion on the regional thermal and humidity segmented dataset to generate a regional cleaning thermal and humidity dataset.

[0074] Specifically, temperature, humidity, zone heating status, local sweating status, and joint motion status information corresponding to adjacent phase windows within the jurisdiction of the same edge node are extracted from the regional thermal and humidity segmentation dataset. Based on the temporal sequence, state succession, and trend correspondence between the end position of the previous phase window and the beginning position of the next phase window, a before-and-after comparison is performed on the temperature, humidity, zone heating status, local sweating status, and joint motion status information to form a before-and-after comparison result. Based on the before-and-after comparison result, temperature shifts, humidity shifts, zone heating status shifts, local sweating status shifts, and joint motion status shifts between adjacent phase windows are analyzed. The system performs drift correction on joint motion state offset, time correction on temporal misalignment between adjacent phase windows, and anomaly identification on temperature, humidity, zone heating state, local sweating state, and joint motion state information that exhibit state jumps, change reversals, time sequence inversions, discontinuities, or contradictions. It also fills in missing positions according to the correspondence between adjacent phase windows, ensuring that the temperature, humidity, zone heating state, local sweating state, and joint motion state information remain continuously corresponding between adjacent phase windows after drift correction, time correction, anomaly identification, and missing information filling, thus generating a regional cleaning thermal and humidity dataset.

[0075] S3.3. Standardize the format and arrange the order of the regional cleaning thermal and humidity dataset to generate regional preprocessed thermal and humidity data.

[0076] Specifically, based on the regional cleaning thermal and humidity dataset, the recording formats, time stamp formats, and phase stamp formats for temperature information, humidity information, zone heating status information, local sweating status information, and joint movement status information are uniformly organized. Following the jurisdiction of edge nodes, the order of phase stamps, and the chronological order, the uniformly formatted temperature information, humidity information, zone heating status information, local sweating status information, and joint movement status information are continuously arranged and correspondingly merged. This ensures that the temperature information, humidity information, zone heating status information, local sweating status information, and joint movement status information form a coherent and consistently labeled thermal and humidity time series within the same edge node jurisdiction, generating regional preprocessed thermal and humidity data.

[0077] S3.4 Extract the change trajectories of temperature, humidity, zone heating status, local sweating status and joint movement status within the continuous phase window from the regional preprocessed thermal and humidity data, and perform correlation analysis to generate local thermal and humidity change patterns.

[0078] Specifically, temperature, humidity, zoned heating status, localized sweating status, and joint motion status information corresponding to continuous phase windows within the jurisdiction of the same edge node are extracted from the preprocessed thermal and humidity data of the region. These information are then continuously read according to phase sequence and chronological order, forming the trajectory of each type of information within the continuous phase window. Based on the relationship between the changes in temperature, humidity, zoned heating status, localized sweating status, and joint motion status information in adjacent phase windows, the relationship between temperature changes and humidity changes, and between zoned heating status changes and localized sweating status changes, is analyzed chronologically. The changes in sweat state and joint movement state are analyzed in correspondence with changes in temperature and humidity. Relationships showing sequential correspondence, continuity of change processes, and mutual response among temperature and humidity information, zoned heating state information and localized sweating state information, and joint movement state information and changes in temperature and humidity information are identified as correlated relationships. Relationships showing temporal discontinuity, interrupted change, conflicting direction, or lack of corresponding response among temperature and humidity information, zoned heating state information and localized sweating state information, and joint movement state information and changes in temperature and humidity information are identified as uncorrelated relationships. Based on the correlated relationships, the sequential and corresponding relationships of various information types within the continuous phase window are integrated to generate local temperature and humidity change patterns.

[0079] S3.5. Based on the local heat and humidity variation patterns and the preprocessed heat and humidity data of the region, predict the temperature information, humidity information, zone heating status information, local sweating status information and joint movement status information in the next phase window, and generate a set of regional heat and humidity prediction results.

[0080] Specifically, based on local heat and humidity variation patterns and pre-processed regional heat and humidity data, the temperature, humidity, zone heating status, local sweating status, and joint movement status information within the current phase window are read according to the jurisdiction of edge nodes, phase marking order, and time sequence. Based on the established sequence and correspondence of changes in the local heat and humidity variation patterns, the corresponding changes in temperature and humidity, zone heating and local sweating, and joint movement and heat and humidity variation information within the current phase window are analyzed item by item. The direction, sequence, and corresponding status of each type of information within the next phase window are then calculated. Finally, the calculated temperature, humidity, zone heating, local sweating, and joint movement information within the next phase window are organized according to the jurisdiction of edge nodes, phase marking order, and time sequence to generate a regional heat and humidity prediction result set.

[0081] S3.6 Match and compare the regional thermal and humidity prediction result set with the regional preprocessed thermal and humidity data to generate a regional thermal and humidity prediction comparison set.

[0082] Specifically, the regional heat and humidity prediction result set and the regional pre-processed heat and humidity data are arranged in a corresponding order according to the jurisdiction of the edge nodes, the phase label order, and the time sequence. Then, the temperature information, humidity information, zone heating status information, local sweating status information, and joint movement status information are matched and the differences are recorded item by item, so that the predicted content and the actual content of the regional heat and humidity prediction result set and the regional pre-processed heat and humidity data correspond to each other, and a regional heat and humidity prediction comparison set is generated.

[0083] It should be noted that the edge nodes have the ability to understand the time sequence of local thermal and humidity evolution processes. They can integrate temperature information, humidity information, zone heating status information, local sweating status information, and joint movement status information into the same prediction and comparison framework, thereby forming a continuous and discriminative regional difference representation. This provides a more stable data foundation and more timely edge response capability for subsequent edge deviation identification, segment hierarchical processing, edge residual feature encapsulation, and whole-body thermal and humidity state reconstruction.

[0084] S4. Based on the regional thermal and humidity prediction reference set, identify edge deviations and perform segment grading to generate edge residual feature packages.

[0085] S4.1 Extract prediction difference information and temporal change information from the regional heat and humidity prediction control set, and aggregate them according to the jurisdiction of the edge nodes to generate a regional deviation analysis dataset.

[0086] Specifically, the predicted and actual data for each edge node's jurisdiction within a continuous phase window are retrieved from the regional thermal and humidity prediction reference set. The differences between the predicted and actual data are then merged and organized according to the edge node's jurisdiction, phase marker order, and chronological order to form prediction difference information. Based on the changes in the position, sequence, and continuity of these prediction difference information within the continuous phase window, the prediction difference information is continuously read and correlated to form time-series change information. Finally, the prediction difference information and time-series change information are grouped according to the edge node's jurisdiction, ensuring that they form coherent deviation analysis content within the same edge node's jurisdiction, thus generating a regional deviation analysis dataset.

[0087] S4.2 Based on the regional deviation analysis dataset, perform continuity analysis and deviation intensity analysis on the prediction difference information and temporal change information within the jurisdiction of each edge node, identify edge deviation situations and mark corresponding segments, and generate a regional segment hierarchical dataset.

[0088] Specifically, predicted difference information and temporal change information are read from the regional deviation analysis dataset according to the jurisdiction of edge nodes, the order of phase marking, and the chronological order. Correspondence analysis is performed on adjacent predicted difference information and temporal change information within continuous phase windows. The number of phase windows with the same deviation type appearing consecutively at adjacent time positions is determined as the duration, the start and end positions of consecutive appearances are determined as the continuation positions, and the degree of consistency in change direction, continuity, and positional correspondence is determined as the degree of change coherence, forming a continuity analysis result. Based on the magnitude, amplitude, and degree of clustering of predicted difference information at each phase and time position, deviation intensity analysis is performed on predicted difference information and temporal change information, and the edge deviation index of each edge node's jurisdiction within each phase window is calculated, forming the deviation intensity analysis result, expressed as:

[0089] ;

[0090] in, Indicates the first The jurisdiction of the edge node is in the first Edge deviation index within a phase window Indicates the first The jurisdiction of the edge node is in the first The intensity of predicted difference information within each phase window Indicates the first The jurisdiction of the edge node is in the first The continuity intensity of temporal variation information within each phase window Indicates the first The jurisdiction of the edge node is in the first Phase response difference intensity within each phase window Indicates the number of the area under the jurisdiction of the edge node. Indicates the phase window number, This represents the weighting coefficient of the intensity of predicted difference information in the marginal deviation index. This represents the weighting coefficient of the intensity of continuity of temporal change information in the marginal deviation index. This represents the weighting coefficient of the phase response difference intensity in the marginal deviation index. For example, the weighting coefficient of the prediction difference information intensity in the marginal deviation index is 0.4 to 0.6, the weighting coefficient of the temporal change information continuity intensity is 0.2 to 0.4, and the weighting coefficient of the phase response difference intensity is 0.1 to 0.3.

[0091] and When the difference between the predicted content and the actual content plays a major role in edge deviation identification, the weight coefficient corresponding to the intensity of the predicted difference information is increased; when the continuity relationship between the preceding and following phases in the continuous phase window plays a major role in deviation identification, the weight coefficient corresponding to the intensity of the continuity of temporal change information is increased; when the difference in phase switching response plays a supplementary and distinguishing role in deviation identification, the weight coefficient corresponding to the intensity of phase response difference is decreased.

[0092] The continuity analysis results and deviation intensity analysis results are mapped to each segment within the continuous phase window. Based on the magnitude of the edge deviation index, the edge deviation of each segment is identified and the corresponding segment is marked. This allows the prediction difference information, temporal change information, continuity analysis results, and deviation intensity analysis results to form hierarchical content with sequential correlation within the same edge node jurisdiction area, generating a regional segment hierarchical dataset.

[0093] S4.3. Based on the regional segment hierarchical dataset, aggregate and encapsulate the prediction difference information, temporal change information, phase labeling information and regional identification information of different hierarchical segments to generate edge residual feature packages.

[0094] Specifically, prediction difference information, temporal change information, phase label information, and regional identification information corresponding to different graded segments are read from the regional segment hierarchical dataset according to the jurisdiction of edge nodes, the order of phase labels, and the chronological order. Prediction difference information, temporal change information, phase label information, and regional identification information within the same graded segment are then organized accordingly. Based on the sequential positional relationship and hierarchical affiliation of different graded segments within the jurisdiction of edge nodes, prediction difference information, temporal change information, phase label information, and regional identification information are continuously merged and combined, forming a one-to-one segment feature content within the same graded segment. The combined and encapsulated segment feature content is then aggregated according to the jurisdiction of edge nodes, forming continuously readable edge feature content based on the prediction difference information, temporal change information, phase label information, and regional identification information corresponding to different graded segments, generating an edge residual feature package.

[0095] It should also be noted that different graded segments refer to dividing the corresponding segments within a continuous phase window into low deviation segments, medium deviation segments, and high deviation segments based on the magnitude of the difference in the predicted difference information, the duration of the temporal change information, and the strength of the change.

[0096] The regional identification information comes from the regional number or regional location identification content formed after establishing the correspondence between each anatomical region and the jurisdiction of the edge node in the regional distribution data.

[0097] S5. Based on the edge residual feature package, fuse the temporal information of each region and reconstruct the whole-body thermal and humidity state to generate the whole-body thermal and humidity fusion result.

[0098] S5.1 Based on the phase labeling information and region identification information, the edge residual feature package is time-series aggregated to generate a regional time-series fusion dataset.

[0099] Specifically, prediction difference information, temporal change information, phase labeling information, and region identification information are read from the edge residual feature package. These information are then categorized by region identification information. Next, based on phase labeling information and temporal order, the prediction difference information, temporal change information, phase labeling information, and region identification information corresponding to the same region identification information are sorted sequentially and merged continuously, ensuring a coherent temporal arrangement under the same region identification information. Finally, the categorized and temporally arranged prediction difference information, temporal change information, phase labeling information, and region identification information are organized to generate a regional temporal fusion dataset.

[0100] S5.2 Perform cross-regional alignment and continuous stitching on the regional temporal fusion dataset to generate a whole-body thermal and humidity reconstruction dataset.

[0101] Specifically, based on the regional temporal fusion dataset, the predicted difference information, temporal change information, and phase label information corresponding to different regional identifiers are synchronized according to phase label information and temporal order. Based on the adjacency and phase correspondence of each regional identifier information in the whole-body distribution, the predicted difference information, temporal change information, and phase label information that have completed synchronization are aligned across regions, ensuring that the predicted difference information, temporal change information, and phase label information corresponding to different regional identifiers form an adjacency relationship at the same phase and time positions. The cross-regionally aligned predicted difference information, temporal change information, and phase label information are then continuously spliced ​​according to the whole-body position order and temporal continuity order, forming a complete and coherent whole-body thermo-humidity temporal content, generating a whole-body thermo-humidity reconstruction dataset.

[0102] S5.3 Perform overall association and state reconstruction on the whole-body thermal and moisture reconstruction dataset to generate whole-body thermal and moisture fusion results.

[0103] Specifically, the predicted difference information, temporal change information, and phase marker information corresponding to the regional identifier information are read from the whole-body thermal and moisture reconstruction dataset. These information are then systematically organized according to the whole-body position order, phase marker order, and chronological order. Based on the sequential changes of different regional identifier information at the same phase position and adjacent time positions, the predicted difference information and temporal change information are correlated as a whole, forming a continuous whole-body thermal and moisture change content. Following this correlation, the predicted difference information, temporal change information, and phase marker information corresponding to the regional identifier information within the whole-body location range are read sequentially. Based on the whole-body position order, phase marker order, and chronological order, continuously correlated thermal and moisture change content is merged, while discontinuous content is supplemented and rearranged to reconstruct the thermal and moisture change state corresponding to each regional identifier information. This ensures that the predicted difference information, temporal change information, and phase marker information remain consistent and coherent throughout the whole-body location range, generating a whole-body thermal and moisture fusion result.

[0104] S6. Correct the acquisition strategy and update the data back to the whole body thermo-wet fusion results to generate an updated gait phase edge acquisition reference set.

[0105] S6.1 Extract thermal and moisture deviation distribution information, phase response difference information, and temporal change intensity information from the whole-body thermal and moisture fusion results for different edge node jurisdiction areas, and collect them according to regional identification information and phase label information to generate a set of acquisition strategy correction information.

[0106] Specifically, the thermal and moisture change status content corresponding to the regional identification information is read from the whole-body thermal and moisture fusion results. According to the correspondence between the edge node's jurisdiction area and the regional identification information, the thermal and moisture change status content corresponding to the regional identification information is merged into regions to form thermal and moisture deviation distribution information, phase response difference information, and temporal change intensity information corresponding to different edge node jurisdiction areas. Based on the regional identification information and phase marker information, the thermal and moisture deviation distribution information, phase response difference information, and temporal change intensity information are sorted and correspondingly collected to generate a data acquisition strategy correction information set.

[0107] S6.2. Based on the acquisition strategy, modify the information set, adjust the sampling start time, phase marking rules and upload trigger conditions of the jurisdiction of each edge node, and generate an edge acquisition update information set.

[0108] Specifically, the information set is modified according to the acquisition strategy. Based on the correspondence between the jurisdiction of the edge nodes and the area identification information, the thermal and humidity deviation distribution information, phase response difference information, and temporal change intensity information are organized into regional correspondences. According to the distribution of thermal and humidity deviation distribution information in different edge node jurisdictions, the sampling start time of each edge node jurisdiction is advanced, delayed, or maintained. According to the change relationship of phase response difference information in adjacent phase positions, the phase marking rules of each edge node jurisdiction are refined and adjusted. According to the strength of change of temporal change intensity information in continuous time positions, the upload trigger conditions of each edge node jurisdiction are increased, decreased, or reset. The adjusted sampling start time, phase marking rules, and upload trigger conditions are then aggregated according to the edge node jurisdiction to generate an edge acquisition update information set.

[0109] It should also be noted that the refinement adjustment refers to further distinguishing the phase division position, phase switching judgment content, and phase sequence corresponding content based on the original phase marking rules; when the phase response difference information is continuously concentrated near the original phase division boundary, or when the same phase marking content corresponds to multiple different response situations, the original phase interval is split into smaller phase intervals, or transition judgment content between adjacent phases is added, as a refinement adjustment of the phase marking rules; when the intensity of time sequence change information continuously increases in continuous time position, the changes before and after appear in concentrated form, or the upload lags behind the actual change, the upload trigger condition is increased; when the intensity of time sequence change information continuously weakens in continuous time position, the changes before and after tend to stabilize, or the upload frequency is higher than the change requirement, the upload trigger condition is decreased; when the original upload trigger condition can no longer correspond to the current time sequence change process, or when the change order, change range, and trigger basis in continuous time position all change, the upload judgment content is redefined, as a reset of the upload trigger condition.

[0110] S6.3. The edge acquisition update information set is transmitted back to the corresponding edge node jurisdiction area, and the gait phase edge acquisition reference set is updated to generate the updated gait phase edge acquisition reference set.

[0111] Specifically, based on the correspondence between the edge node's jurisdiction area and the edge acquisition update information set, the sampling start time, phase marking rules, and upload trigger conditions in the edge acquisition update information set are sent to the corresponding edge node's jurisdiction area. The original sampling start time, sampling duration, phase marking rules, and upload trigger conditions of the corresponding edge node's jurisdiction area are read from the gait phase edge acquisition reference set. The sampling start time, phase marking rules, and upload trigger conditions in the edge acquisition update information set are then replaced and updated sequentially with their corresponding contents in the gait phase edge acquisition reference set, while retaining the original sampling duration of the corresponding edge node's jurisdiction area. According to the edge node's jurisdiction area, phase marking order, and time sequence, the updated sampling start time, sampling duration, phase marking rules, and upload trigger conditions are correspondingly organized and continuously arranged, ensuring that the edge acquisition update information set and the gait phase edge acquisition reference set form a seamless update result within the same edge node's jurisdiction area, generating the updated gait phase edge acquisition reference set.

[0112] This embodiment also provides a computer device applicable to the edge computing-based multi-channel thermal and humidity acquisition method, comprising: a memory and a processor; the memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions to implement the edge computing-based multi-channel thermal and humidity acquisition method proposed in the above embodiment.

[0113] The computer device can be a terminal, comprising a processor, memory, communication interface, display screen, and input devices connected via a system bus. The processor provides computing and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system and computer programs. The internal memory provides an environment for the operation of the operating system and computer programs stored in the non-volatile storage media. The communication interface is used for wired or wireless communication with external terminals; wireless communication can be achieved through Wi-Fi, carrier networks, NFC (Near Field Communication), or other technologies. The display screen can be an LCD screen or an e-ink screen. The input devices can be a touch layer covering the display screen, buttons, a trackball, or a touchpad on the computer device's casing, or an external keyboard, touchpad, or mouse.

[0114] This embodiment also provides a storage medium storing a computer program that, when executed by a processor, implements the edge computing-based multi-channel thermal and humidity acquisition method as proposed in the above embodiments. The storage medium can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as Static Random Access Memory (SRAM), Electrically Erasable Programmable Read-Only Memory (EEPROM), Erasable Programmable Read Only Memory (EPROM), Programmable Red-Only Memory (PROM), Read-Only Memory (ROM), magnetic storage, flash memory, magnetic disk, or optical disk.

[0115] In summary, this invention identifies edge deviations by using a regional thermal and humidity prediction control set and performs segment-level processing to generate edge residual feature packages. This enables the continuous thermal and humidity change process to be transformed into a structured representation with deviation intensity, temporal location, and regional affiliation at the edge computing node. This allows for the centralized expression of the coupling features between local temperature and humidity changes, regional heating responses, local sweating responses, and motion state perturbations, forming a high-information-density data carrier for subsequent cross-regional temporal fusion. This improves the temporal coherence, regional identification, and update responsiveness in the whole-body thermal and humidity state reconstruction process.

[0116] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

Claims

1. A multi-channel thermal and humidity acquisition method based on edge computing, characterized in that: include, Collect regional distribution data and motion state data, and perform edge node deployment and unified timing processing to generate a gait phase edge acquisition reference set; Phase-locked synchronous acquisition and phase alignment processing are performed on the gait phase edge acquisition reference set to generate phase-locked thermal and humidity raw data stream; Edge preprocessing is performed on the phase-locked thermal and humidity raw data stream to generate regional preprocessed thermal and humidity data, and local thermal and humidity variation patterns are extracted for time series extrapolation and comparison verification to generate a regional thermal and humidity prediction comparison set; Based on the regional thermal and humidity prediction reference set, edge deviation is identified and segment grading is performed to generate edge residual feature packages. Based on the edge residual feature packets, the temporal information of each region is fused and the whole-body thermal and humidity state is reconstructed to generate the whole-body thermal and humidity fusion result. The acquisition strategy of the whole-body thermo-moisture fusion results is corrected and updated by backhaul, and an updated gait phase edge acquisition reference set is generated.

2. The edge computing-based multi-channel thermal and humidity acquisition method as described in claim 1, characterized in that: The specific steps for generating the gait phase edge acquisition reference set are as follows: Information on the anatomical boundaries, zoned heating distribution, localized sweating distribution, and joint movement trajectories of the warm-body dummy were collected and organized according to a unified time reference to generate the original regional motion dataset. Based on the original regional motion dataset, spatial correspondence and temporal association are performed on the anatomical region boundary information, zoned heating distribution information, local sweating distribution information and joint motion trajectory information to generate regional distribution data and motion state data. Based on regional distribution data and motion status data, the jurisdiction of edge nodes is divided and access relationships are established. Unified timing is performed in combination with the order of attitude switching, and the edge node deployment timing result is generated. Based on the timing results of edge node deployment, the corresponding sampling start time, sampling duration, phase marking rules and upload trigger conditions are configured for the jurisdiction of each edge node, generating a gait phase edge acquisition reference set.

3. The edge computing-based multi-channel thermal and humidity acquisition method as described in claim 2, characterized in that: The specific steps for generating the phase-locked thermal and humidity raw data stream are as follows. Based on the gait phase edge acquisition reference set, the sampling start time, sampling duration, phase marking rules and upload trigger conditions of the area under the jurisdiction of each edge node are sent to the corresponding edge node to generate a phase-locked acquisition control set; According to the phase-locked acquisition control set, each edge node synchronously acquires temperature information, humidity information, zone heating status information, local sweating status information and joint motion status information within the corresponding phase window, and performs edge association to generate a regional phase-locked acquisition segment set. Based on the regional phase-locked acquisition fragment set, the temperature information, humidity information, zone heating status information, local sweating status information and joint movement status information within the jurisdiction of each edge node are corrected for local time offset according to the phase marking rules, and then spliced ​​and organized according to a unified time order and phase sequence to generate the phase-locked thermal and humidity raw data stream.

4. The edge computing-based multi-channel thermal and humidity acquisition method as described in claim 3, characterized in that: The specific steps for preprocessing the thermal and humidity data of the generated region are as follows. By utilizing the jurisdiction of edge nodes and phase marking rules, the original phase-locked thermal and humidity data stream is segmented and correlated to generate a regional thermal and humidity segmented dataset; Based on the continuity of change between adjacent phase windows, drift correction, time correction, anomaly identification and missing completion are performed on the regional thermal and humidity segmented dataset to generate a regional cleaning thermal and humidity dataset. The regional cleaning thermal and humidity dataset is standardized in format and ordered to generate regional preprocessed thermal and humidity data.

5. The edge computing-based multi-channel thermal and humidity acquisition method as described in claim 4, characterized in that: The specific steps for generating the regional thermal and humidity prediction control set are as follows. The change trajectories of temperature, humidity, zone heating status, local sweating status and joint movement status within a continuous phase window are extracted from the preprocessed regional thermal and humidity data, and correlation analysis is performed to generate local thermal and humidity change patterns. Based on the local heat and humidity variation patterns and the preprocessed heat and humidity data of the region, the temperature information, humidity information, zone heating status information, local sweating status information and joint movement status information in the next phase window are predicted to generate a set of regional heat and humidity prediction results. The regional thermal and humidity prediction result set is matched and compared with the regional preprocessed thermal and humidity data to generate a regional thermal and humidity prediction comparison set.

6. The edge computing-based multi-channel thermal and humidity acquisition method as described in claim 5, characterized in that: The specific steps for generating the edge residual feature packet are as follows: Prediction difference information and temporal change information are extracted from the regional heat and humidity prediction control set, and then aggregated according to the jurisdiction of the edge nodes to generate a regional deviation analysis dataset. Based on the regional deviation analysis dataset, we perform continuity analysis and deviation intensity analysis on the prediction difference information and temporal change information within the jurisdiction of each edge node, identify edge deviation situations and label corresponding segments, and generate a regional segment hierarchical dataset. Based on the regional segment hierarchical dataset, the prediction difference information, temporal change information, phase labeling information and regional identification information of different hierarchical segments are aggregated and encapsulated to generate edge residual feature packages.

7. The edge computing-based multi-channel thermal and humidity acquisition method as described in claim 6, characterized in that: The specific steps for generating the whole-body thermo-moisture fusion result are as follows. Based on phase marker information and region identifier information, the edge residual feature packets are temporally aggregated to generate a regional temporal fusion dataset. Cross-regional alignment and continuous stitching are performed on the regional temporal fusion dataset to generate a whole-body thermal and humidity reconstruction dataset; The whole-body thermo-humidity reconstruction dataset is correlated and its state is reconstructed to generate the whole-body thermo-humidity fusion result.

8. The edge computing-based multi-channel thermal and humidity acquisition method as described in claim 7, characterized in that: The specific steps for generating the updated gait phase edge acquisition reference set are as follows: The thermal and moisture deviation distribution information, phase response difference information, and temporal change intensity information of different edge node jurisdiction areas are extracted from the whole-body thermal and moisture fusion results, and then collected according to the regional identification information and phase label information to generate a set of acquisition strategy correction information. Based on the acquisition strategy, the information set is modified, and the sampling start time, phase marking rules, and upload triggering conditions of the jurisdiction of each edge node are adjusted to generate an edge acquisition update information set. The edge acquisition update information set is sent back to the corresponding edge node jurisdiction area, and the gait phase edge acquisition reference set is updated to generate the updated gait phase edge acquisition reference set.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that: When the processor executes the computer program, it implements the steps of the edge computing-based multi-channel thermal and humidity acquisition method according to any one of claims 1 to 8.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by the processor, it implements the steps of the edge computing-based multi-channel thermal and humidity acquisition method according to any one of claims 1 to 8.

Images