A distributed indoor temperature and humidity data processing system and method

Abstract

This invention relates to the field of environmental management technology, specifically a distributed indoor temperature and humidity data processing system and method. The system includes a node data initialization module, a state difference extraction module, a multiple offset determination module, a fluctuation priority labeling module, and a data processing scheduling module. In this invention, by performing paired difference analysis on the temperature and humidity sampling values ​​of multiple nodes in an indoor area and establishing a binding structure, a state difference ratio sequence is constructed by combining historical difference averages to extract key fluctuation indicators, achieving dynamic identification of state changes within a period. By comparing the offset ratio with the response threshold, nodes in areas with significant fluctuations are screened, clearly defining the areas to be processed in the spatial distribution dimension and forming a map basis. By comparing the data average of consecutive periods with a sensitive threshold, high-priority and ordinary-priority data units are distinguished, achieving hierarchical scheduling and time-series processing, improving the accuracy of identifying and controlling environmental state anomalies.

Description

Technical Field

[0001] This invention relates to the field of environmental management technology, and in particular to a distributed indoor temperature and humidity data processing system and method. Background Technology

[0002] The field of environmental management technology involves the monitoring, control, and optimization of environmental parameters within residential or work spaces. This includes the data collection, transmission, and analysis of key environmental elements such as temperature, humidity, air quality, light intensity, and noise. This data supports comprehensive management objectives such as building energy consumption management, comfort adjustment, and safety control. The system systematically encompasses sensor network deployment, environmental data management processes, and control strategy execution, forming an environmental information processing framework integrating sensing, computing, and scheduling. Specifically, an indoor temperature and humidity data processing system is used to collect and process indoor temperature and humidity information to assist in environmental regulation. The technical challenge it addresses is how to efficiently collect, centrally process, and determine the status of distributed temperature and humidity data across multiple spatial areas. Typically, fixed-point sensors are connected to a central processing unit via a bus. The collected data is transmitted to the main control unit using temperature and humidity acquisition circuits and serial communication protocols. The main control unit then uses threshold comparison methods to determine and record the temperature and humidity status.

[0003] In the current process of collecting indoor temperature and humidity data, fixed-point sensors connected in a centralized manner via a bus mainly transmit the collected data to the main control device for judgment and processing. In scenarios with multiple concurrent nodes or wide spatial distribution, the centralized collection path and the single judgment method can easily lead to data update lag. Furthermore, the use of threshold comparison methods cannot effectively identify the state trend deviation caused by periodic subtle changes. If some nodes gradually deviate from the normal state within a continuous period but do not cross the fixed threshold, they may not be identified and recorded in time, affecting the timeliness and accuracy of subsequent environmental adjustment decisions. In residential or work spaces with high requirements for fluctuation sensitivity, the support capability for actual comfort and energy efficiency control is significantly insufficient. Summary of the Invention

[0004] The purpose of this invention is to address the shortcomings of existing technologies by proposing a distributed indoor temperature and humidity data processing system and method.

[0005] To achieve the above objectives, the present invention adopts the following technical solution: a distributed indoor temperature and humidity data processing system comprising:

[0006] The node data initialization module obtains the node number of the indoor area, collects the temperature and humidity sampling values ​​of the current period, establishes a one-to-one corresponding data set based on the node number, and binds the node number pair with the corresponding temperature and humidity data to generate the node difference initialization dataset.

[0007] The state difference extraction module initializes the dataset based on the node difference, calculates the difference between the temperature and humidity sampling values ​​of each pair of nodes, binds them to the node number pairs in sequence to form difference record units, performs difference calculation between the difference record units and the average difference of the node's historical three periods, extracts the maximum value of the temperature difference ratio and humidity difference ratio, and obtains the state difference index sequence between nodes.

[0008] The multiple offset determination module filters out all node pairs whose maximum ratio value exceeds the response threshold multiple based on the state difference index sequence between nodes, and establishes a corresponding regional number mapping to obtain the distribution map of multiple offset nodes within the period.

[0009] The fluctuation priority labeling module calculates the average temperature and humidity data of each node in the current cycle and the previous two cycles based on the distribution map of the multiple offset nodes within the cycle. If the average temperature and humidity data exceeds the fluctuation sensitivity threshold, the node is marked as a high-priority data unit; otherwise, it is marked as a normal-priority unit, thus generating a set of node-level temperature and humidity fluctuation priority identifiers.

[0010] As a further embodiment of the present invention, the node difference initialization dataset includes a node number pair mapping relationship, a temperature and humidity sampling dataset within a period, and a data set binding structure; the state difference index sequence includes a temperature difference ratio sequence between nodes, a humidity difference ratio sequence between nodes, and a record item of the maximum value of the ratio difference; the node distribution map of multiple offset within a period includes node pair numbers exceeding the threshold, a node region number mapping relationship, and an abnormal distribution map structure within a period; the node-level temperature and humidity fluctuation priority identifier set includes the node temperature and humidity average, the comparison result with the fluctuation sensitivity threshold, and the priority identifier label.

[0011] As a further aspect of the present invention, the node data initialization module includes:

[0012] The node number extraction submodule obtains all adjacent node numbers from the indoor area node number table, identifies each pair of node number pairs with a direct connection relationship, and establishes a set of number pairs to generate a node number pair set.

[0013] The environmental data acquisition submodule reads the temperature and humidity sampling values ​​corresponding to each pair of node numbers in the current period based on the set of node number pairs, extracts the environmental sampling values ​​of each node, and combines them to construct a double set of environmental datasets under the node pair, thus obtaining a double set of environmental sampling values.

[0014] The difference data generation submodule extracts the sequence number, period number and environmental data combination information of each pair of nodes based on the dual set of environmental sample values, organizes them in a unified structure, binds each node number pair to its corresponding environmental data set, and generates the node difference initialization dataset.

[0015] As a further aspect of the present invention, the state difference extraction module includes:

[0016] The difference record generation submodule initializes the dataset based on the node difference, extracts the current cycle temperature and humidity sampling values ​​corresponding to each pair of node numbers, establishes a set of difference data for the temperature difference and humidity difference values ​​of each pair of nodes, unifies the format of all records and includes them in the cache set, and establishes node environment difference records.

[0017] The periodic difference comparison submodule, based on the node environmental difference record, uses the node number pair and the current period number as index keys to read the historical data of temperature difference and humidity difference values ​​of the node number pair in the previous three consecutive periods. It then processes the difference ratio between the current period temperature difference value and the average temperature difference value of the previous three periods, and performs the same operation on the current period humidity difference value and the average humidity difference value of the previous three periods. The periodic temperature difference ratio and periodic humidity ratio are calculated and obtained respectively, and then uniformly bound to the node number pair and the period number to obtain the node difference ratio sequence.

[0018] The difference index extraction submodule extracts the temperature difference ratio and humidity difference ratio records of each node pair in the current cycle based on the node difference ratio sequence, extracts the maximum value and binds it with the node number to form a record structure, and stores it in the index sequence set to establish a state difference index sequence.

[0019] As a further aspect of the present invention, the multiple offset determination module includes:

[0020] The threshold comparison submodule extracts the temperature difference ratio and humidity difference ratio of the current period from the node difference ratio sequence based on the temperature difference ratio and humidity difference ratio in the inter-node state difference index sequence, calculates and obtains the offset exceeding the threshold intensity value, and filters out node pairs whose offset exceeds the trend based on whether the intensity value is greater than zero, and generates a set of offset exceeding the limit node numbers.

[0021] The region mapping submodule divides the regions according to the number prefix of the node number pair based on the offset over-limit node number set, extracts the region identifier carried in the number as the region mapping index key, and clusters all over-limit node pairs according to the region number to obtain the region offset node index structure.

[0022] The node graph generation submodule obtains the set of node number pairs under each region according to the region offset node index structure, generates node connection edges according to the region number as the index, constructs an undirected graph structure according to the order of node pairs, registers the graph structure with the period as the index number, and generates a node distribution map of multiple offset within the period.

[0023] As a further aspect of the present invention, the fluctuation priority labeling module includes:

[0024] The periodic data extraction submodule extracts the raw temperature and humidity data uploaded by each node in the current period and the previous two periods based on the node number in the distribution map of the multiple offset nodes within the period. The data is then categorized and organized according to the node number, and a temperature and humidity data sequence indexed by the period is generated according to the node number to establish a three-period temperature and humidity data structure.

[0025] The threshold judgment submodule reads the temperature and humidity data corresponding to the current cycle in terms of nodes according to the three-cycle temperature and humidity data structure, compares it with the same type of data in the previous two cycles, performs difference level confirmation operation for each type of data, sets temperature and humidity fluctuation judgment standard values, determines whether it exceeds the perceptible range, identifies and archives the node priority status, and obtains the node fluctuation status label set.

[0026] The priority identifier generation submodule extracts the node number and status identifier content marked as fluctuating in the current cycle based on the node fluctuation status label set, and establishes a time index in combination with the cycle number. It then outputs the status result information of each node in a structured manner, maps and integrates the node number and priority identifier, and establishes a node-level temperature and humidity fluctuation priority identifier set.

[0027] As a further aspect of the present invention, the system further includes:

[0028] The data processing and scheduling module inserts data into the priority scheduling processing pool and the normal scheduling processing pool according to the priority type based on the node-level temperature and humidity fluctuation priority identifier set, and executes the data processing sequentially according to the time order, thereby establishing a distributed indoor temperature and humidity data processing record table.

[0029] The distributed indoor temperature and humidity data processing record table includes node number record items, scheduling time point information, processing resource type identifier, and task execution order record items.

[0030] As a further aspect of the present invention, the data processing scheduling module includes:

[0031] The priority pool partitioning submodule, based on the priority tag information of all nodes in the node-level temperature and humidity fluctuation priority identifier set, writes the data unit corresponding to the high-priority node number into the priority scheduling processing pool and writes the data unit corresponding to the ordinary priority node number into the ordinary scheduling processing pool. It sorts the data in ascending order according to the timestamp carried by the node number, removes data entries with time conflicts, and caches the data sets of each priority type in a queue structure to generate a priority scheduling dataset.

[0032] The time queue management submodule sets up two scheduling processing channels based on the priority scheduling dataset and the scheduling priority type. In each scheduling cycle, it performs dequeue processing on the data unit at the head of the queue. Each processing records the current system timestamp and uses the time as the scheduling time point of the data unit. It records the dequeue sequence number and organizes the records with the node number as the primary key to obtain the data processing scheduling sequence record table.

[0033] The scheduling record generation submodule, based on the node number, scheduling time point, processing resource type, and scheduling sequence number information of each record in the data processing scheduling sequence record table, arranges all fields in ascending order according to the node number, and uses the cycle number and processing time point to generate a unique primary key index to distinguish the task scheduling behavior at different time points in the same cycle, and establishes a distributed indoor temperature and humidity data processing record table.

[0034] A distributed indoor temperature and humidity data processing method includes the following steps:

[0035] S1: Obtain the node number of the indoor area, collect the temperature and humidity sampling values ​​of the current period respectively, establish a one-to-one corresponding data set according to the node number, bind the node number pair with the corresponding temperature and humidity data, and generate the node difference initialization dataset.

[0036] S2: Based on the node difference initialization dataset, calculate the difference of temperature and humidity sampling values ​​for each pair of nodes, bind them sequentially with node number pairs to form difference record units, perform difference operation between the difference record units and the average difference of the node's historical three periods, extract the maximum value of temperature difference ratio and humidity difference ratio, and obtain the state difference index sequence between nodes.

[0037] S3: Based on the state difference index sequence between nodes, filter all node pairs whose maximum ratio value exceeds the response threshold multiple, and establish a corresponding regional number mapping to obtain the distribution map of nodes with multiple offset within the period.

[0038] S4: Based on the distribution map of nodes with multiple offsets within the cycle, calculate the average temperature and humidity data of each node in the current cycle and the previous two cycles. If it exceeds the fluctuation sensitivity threshold, mark it as a high-priority data unit; otherwise, mark it as a normal-priority unit and generate a set of node-level temperature and humidity fluctuation priority identifiers.

[0039] S5: Based on the node-level temperature and humidity fluctuation priority identifier set, insert them into the priority scheduling processing pool and the normal scheduling processing pool according to priority type, and dequeue them in time order to perform data processing in sequence, and establish a distributed indoor temperature and humidity data processing record table.

[0040] Compared with the prior art, the advantages and positive effects of the present invention are as follows:

[0041] In this invention, by performing paired difference analysis on temperature and humidity sampling values ​​of multiple nodes in an indoor area and establishing a binding structure, and combining historical difference averages to construct a state difference ratio sequence to extract key fluctuation indicators, dynamic identification of state changes within a cycle is achieved. By comparing the offset ratio with the response threshold, nodes in areas with significant fluctuations are screened, and the areas to be processed are clearly defined in the spatial distribution dimension to form a map basis. By comparing the data average of continuous cycles with sensitive thresholds, high-priority and ordinary-priority data units are distinguished, enabling hierarchical scheduling and time-series processing. This ensures that the data processing of nodes with abnormal states has timely response and resource adaptability, improving the identification accuracy and control efficiency of environmental state anomalies in scenarios with wide spatial distribution and frequent state fluctuations. Attached Figure Description

[0042] Figure 1 This is a system flowchart of the present invention;

[0043] Figure 2 This is a flowchart of the node data initialization module of the present invention;

[0044] Figure 3 This is a flowchart of the state difference extraction module of the present invention;

[0045] Figure 4 This is a flowchart of the multiplier offset determination module of the present invention;

[0046] Figure 5 This is a flowchart of the fluctuation priority labeling module of the present invention;

[0047] Figure 6 This is a flowchart of the data processing and scheduling module of the present invention. Detailed Implementation

[0048] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention.

[0049] In the description of this invention, it should be understood that the terms "length," "width," "upper," "lower," "front," "rear," "left," "right," "vertical," "horizontal," "top," "bottom," "inner," and "outer," etc., indicating orientation or positional relationships, are based on the orientation or positional relationships shown in the accompanying drawings and are only for the convenience of describing the invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation, and therefore should not be construed as a limitation of the invention. Furthermore, in the description of this invention, "a plurality of" means two or more, unless otherwise explicitly specified.

[0050] Please see Figure 1A distributed indoor temperature and humidity data processing system includes:

[0051] The node data initialization module obtains the node numbers of each pair of adjacent nodes from the indoor area node number table, collects the temperature and humidity sampling values ​​of the current period, establishes a one-to-one corresponding data set based on the node numbers, and binds each pair of node numbers to the corresponding temperature and humidity data set to generate the node difference initialization dataset.

[0052] The state difference extraction module initializes the dataset based on node differences, calculates the difference between the temperature and humidity sampling values ​​of each pair of nodes, binds the difference results to the node number pairs to form difference record units, performs difference calculation between the difference record units and the average difference of the node's historical three periods, constructs a difference ratio sequence, extracts the maximum value of the temperature difference ratio and humidity difference ratio, and obtains the state difference index sequence between nodes.

[0053] The multiple offset determination module compares the maximum ratio value with the response threshold multiple (the abnormal fluctuation judgment threshold defined by the building environment monitoring standard: 1.8 times the difference of the historical mean) based on the temperature difference ratio and humidity difference ratio of each pair of nodes in the state difference index sequence between nodes. It then filters out all node pairs that exceed the threshold and establishes a corresponding area number mapping to form the node set of the area to be processed, thus obtaining the multiple offset node distribution map within the period.

[0054] The fluctuation priority labeling module extracts the temperature and humidity data uploaded by each node in the current cycle and the previous two cycles based on the node number in the node distribution map of the multiple offset nodes within the cycle. It calculates the average value of each data unit and compares it with the fluctuation sensitivity threshold (the critical value of temperature and humidity changes that can be perceived by the human body as specified in the thermal environment measurement standard, temperature ±0.5℃ / humidity ±3%RH). If the value exceeds the threshold, the data unit is marked as a high-priority data unit; otherwise, it is marked as a normal-priority data unit, thus generating a set of node-level temperature and humidity fluctuation priority identifiers.

[0055] The data processing scheduling module inserts all high-priority and ordinary-priority data units in the node-level temperature and humidity fluctuation priority identifier set into the priority scheduling processing pool (high-priority data processing queue defined by the IoT system architecture standard: response latency ≤100ms) and the ordinary scheduling processing pool (regular processing queue defined by the edge computing standard: response latency ≤500ms) according to priority type. The data is then dequeued in chronological order and processed sequentially. The module records the node number, scheduling time, processing resource type and execution order of each processing task, and establishes a distributed indoor temperature and humidity data processing record table.

[0056] The node difference initialization dataset includes node number mapping relationships, temperature and humidity sampling datasets within the period, and data set binding structure. The state difference index sequence includes the inter-node temperature difference ratio sequence, the inter-node humidity difference ratio sequence, and the record item of the maximum ratio difference. The node distribution map of multiple offset within the period includes node number pairs exceeding the threshold, node region number mapping relationships, and the structure of the abnormal distribution map within the period. The node-level temperature and humidity fluctuation priority identifier set includes the node temperature and humidity average, comparison results with the fluctuation sensitivity threshold, and priority identifier labels. The distributed indoor temperature and humidity data processing record table includes node number record items, scheduling time point information, processing resource type identifier, and task execution order record items.

[0057] Please see Figure 2 The node data initialization module includes:

[0058] The node number extraction submodule obtains all adjacent node numbers from the indoor area node number table, identifies each pair of node number pairs with a direct connection relationship, and establishes a set of number pairs to generate a node number pair set.

[0059] To obtain all adjacent node numbers in the indoor area node numbering table, the node numbering information must first be extracted based on the building floor plan and node deployment planning list. For example, if the node numbers are arranged sequentially from N01 to N20, representing various measuring points located on the first floor, the spatial connectivity between nodes can be determined using the node coordinates on the drawings. The specific standard is that the straight-line distance between nodes is less than or equal to 2 meters, and there are no obstructions in between. This judgment needs to be combined with measurement data. For example, if nodes N01 to N05 are located in passage A1, each 1.5 meters apart, the adjacent relationships can be determined sequentially based on the site deployment diagram and the node coordinate set, identifying node pairs such as N01-N02 and N02-N03. Node pairs that meet the criteria need to be constructed into a numbered pair structure, which can be a two-dimensional array, such as [['N01', 'N02'], ['N02', 'N03'], ['N03', 'N04'], ['N04', 'N05']]. During the identification process, if overlapping numbers or number jumps are found, such as N03-N05 and N03-N04 appearing simultaneously, they need to be screened and cleaned up, retaining the actual physical connection pairs. Node numbers are obtained by synchronizing the sensor initialization log file, filtering and verifying the actual position through the index number field. If the sensor number field is missing or duplicated, it will not be included in the numbered pair set. The final extracted adjacent node pair data is shown in Table 1.

[0060] Table 1 Example data table for node number extraction

[0061]

[0062] As shown in Table 1, the consecutive node pairs formed by N01 to N05 all meet the conditions of 1.5-meter connection distance and no obstruction, and can be added to the numbered set as adjacent nodes to generate a set of node numbered pairs.

[0063] The environmental data acquisition submodule reads the temperature and humidity sampling values ​​corresponding to each pair of node numbers in the current period based on the set of node number pairs, extracts the environmental sampling values ​​of each node, and combines them to construct a double set of environmental datasets under the node pair, thus obtaining a double set of environmental sampling values.

[0064] Based on the node number pair set, it is necessary to further collect the raw temperature and humidity data of the corresponding nodes within the current detection period. Specifically, this involves calling the data cache structure indexed by node number in the data acquisition system, and sequentially reading the environmental sampling values ​​corresponding to each node number. The sampling time is uniformly based on the periodic window of 09:00 on August 8, 2024. In this period, the temperature of node N01 is 24.1℃ and the humidity is 47.8%; the temperature of node N02 is 24.6℃ and the humidity is 49.2%. Therefore, the combination is [24.1, 24.6, 47.8, 49.2], which is the environmental parameter data group for nodes N01-N02 in the current period. All sampling values ​​must meet the condition that the data timestamp difference is less than 20 seconds to be considered data from the same period; otherwise, the data group is discarded to avoid cross-period errors. For example, N... If one of the timestamp records in 03 and N04 is 09:01:02 and the other is 09:02:33, the data will not be counted as valid data for that period because the time difference exceeds the set threshold. After the sampling values ​​are read, they need to be stored in a cache table. The structure is a triple: {node number pair, period number, parameter group array}, for example, {N01-N02, T_2024080809, [24.1, 24.6, 47.8, 49.2]}, where the period number T_2024080809 represents the sampling period at 09:00 on August 8, 2024. Each data storage structure needs to verify the integrity and range of parameters. The temperature needs to be between 15℃ and 35℃, and the humidity needs to be between 30% and 80%. Data outside the range also needs to be removed. All structures that pass the verification constitute a set of two sets of environmental sampling values.

[0065] The difference data generation submodule extracts the sequence number, period number and environmental data combination information of each pair of nodes based on the two sets of environmental sample values, organizes them in a unified structure, binds each node number pair to its corresponding environmental data set, and generates the node difference initialization dataset.

[0066] Based on the two sets of environmental sampling values, the data needs to be structured and organized into a unified dataset. First, each set of data needs to be deconstructed to extract field values ​​such as node number pairs, period number, temperature of node i, temperature of node j, humidity of node i, and humidity of node j, and assigned clear field identifiers. For example, for node pair N01-N02, the four corresponding sampling values ​​in period T_2024080809 are 24.1, 24.6, 47.8, and 49.2, which can be recombined into a six-tuple [N01-N02, T_2024080809, 24.1, 24.6, 47.8, 49.2]. Subsequently, all six-tuples need to be uniformly included in the structured cache set, and a field index table is constructed. The index fields are the node pair and the period number field, which are used for subsequent fast retrieval and matching operations. Before data import, field integrity and format uniformity verification must be performed to ensure that all six fields have been assigned values ​​and that the value types are correct (floating-point numbers). Period archiving processing is performed according to the period number, and all six-tuples are classified and archived by period. Finally, a unified data structure is generated for subsequent modules to perform data difference analysis processing between nodes. After all structured data is imported, it is uniformly stored in the data pool to form the node difference initialization dataset.

[0067] Please see Figure 3 The state difference extraction module includes:

[0068] The difference record generation submodule initializes the dataset based on node differences, extracts the current cycle temperature and humidity sampling values ​​corresponding to each pair of node numbers, establishes a set of difference data for the temperature difference and humidity difference values ​​of each pair of nodes, unifies the format of all records and includes them in the cache set, and establishes node environment difference records.

[0069] To initialize the dataset based on node differences, the first step is to extract each pair of node numbers and their corresponding period numbers. Then, the temperature and humidity samples within the current period are separated from the environmental dataset. The corresponding node data is located using a joint index of the node number and the sampling timestamp. For example, in the period ending at 09:00 on August 8, 2024, if node N01 collects a temperature of 24.8℃ and a humidity of 46.3%, and node N02 collects a temperature of 24.3℃ and a humidity of 48.4%, then the temperature difference between node pairs N01 and N02 in that period is 0.5℃, and the humidity difference is -2.1%. The calculation of the difference... The process involves subtracting node j from node i. The direction of this process is determined by the system's preset node order. The difference structure is uniformly formatted as a four-field value: {node pair number, cycle number, temperature difference value, humidity difference value}. For example, the node pair mentioned above is constructed as {N01-N02, T_2024080809, 0.5, -2.1}. During the recording process, if a node is found to have missing samples or samples that are abnormally outside the range (e.g., temperature outside the 10℃-40℃ range, humidity outside the 10%-90% range), the data for that pair is discarded and removed from the difference record set to ensure data quality. The final structure of the difference record item is as follows:

[0070] Table 2 Example of Node Environment Difference Records

[0071]

[0072] As shown in Table 2, the node difference records for each group in the current period have been organized into a unified structure, which serves as the basic data source for measuring the degree of offset between nodes, and node environment difference records are established.

[0073] The cycle difference comparison submodule, based on the node environmental difference records and using the node number pair and the current cycle number as index keys, reads the historical data of temperature and humidity difference values ​​for the node number pair in the previous three consecutive cycles. It then performs difference ratio processing on the current cycle temperature difference value and the average of the temperature difference values ​​from the previous three cycles, and performs the same operation on the current cycle humidity difference value and the average of the humidity difference values ​​from the previous three cycles, using the following formula:

[0074] ;

[0075] and

[0076] ;

[0077] The periodic temperature difference ratio and periodic humidity difference ratio are calculated separately, and then uniformly bound to the node number pair and the period number to obtain the node difference ratio sequence. Represents node pairs In the cycle The temperature difference, in degrees Celsius. This represents the average temperature difference for that node over the previous three periods, in degrees Celsius. The standard deviation of the temperature difference within the corresponding period is expressed in degrees Celsius. , , These represent the humidity difference for the current period, the mean humidity difference for the previous three periods, and the standard deviation of the humidity difference, respectively, in percentage form. , These represent the temperature difference ratio and the humidity difference ratio, respectively.

[0078] The system retrieves node environment difference records, using the node pair number as the primary index and the period number as the time positioning key. It obtains the original data sequence of temperature and humidity differences for the current period and the three periods preceding it, forming a sequence array structure. The data sampling period is set to once per hour, extracting a total of four groups of samples. The first three groups represent historical periods, and the current group is used to calculate offsets and ratios. For example, if the historical temperature difference sequence for node pairs N01-N02 is [0.4, 0.7, 0.5]℃, and the historical humidity difference is [1.8, 2.0, 2.2]%, and the current period's temperature difference is 0.5℃ and humidity difference is 2.1%, then the average temperature difference is... ℃, mean humidity difference is %; then calculate the standard deviation, temperature standard deviation. ℃, standard deviation of humidity %; Substitute the above calculation results into the normalization formula:

[0079] ;

[0080] ;

[0081] Therefore, the current periodic temperature difference ratio of nodes N01-N02 is 0.1965, and the humidity difference ratio is 0.5, which is used to characterize the degree of difference between the current state and the historical state, and to obtain the node difference ratio sequence.

[0082] The periodic temperature difference ratio and periodic humidity difference ratio are used to measure the degree of deviation of the temperature and humidity differences of nodes in the current detection period from their historical stable state. The periodic temperature difference ratio reflects the intensity of change in the temperature difference between two adjacent nodes in the current period relative to the average temperature difference of the previous three periods. If the ratio is large, it indicates that the temperature distribution in the current period is significantly different from the historical state. Similarly, the periodic humidity difference ratio indicates the degree of deviation of the humidity difference in the current period from the historical average. The higher the ratio, the more unstable the humidity state between nodes. These two ratios are normalized to form corresponding indicators, which are used to uniformly measure the volatility and trend of temperature and humidity in the time evolution process, and provide quantitative support for dynamic monitoring and difference analysis of node status.

[0083] The formula's operational logic is based on the concept of normalized offset. It measures the relative change between the current node state and the historical stable state by standardizing the temperature difference and humidity difference. The numerator is the difference between the current temperature difference (or humidity difference) and the average difference of the previous three periods, reflecting the deviation of the current value from the historical reference level. The denominator is the sample standard deviation of the historical three-period difference, used to measure the fluctuation of historical data. The smaller the standard deviation, the more stable the historical data, and the more significant the impact of the deviation. The entire fractional structure is in the form of a standard fraction, which reflects whether the abnormality of the current state exceeds the historical fluctuation range. The absolute value operation is added to both sides of this ratio to shield the directional influence and uniformly measure the offset intensity. Regardless of whether the change direction is increasing or decreasing, it is considered a valid offset. After independently calculating the temperature difference ratio and humidity difference ratio, they are used as two different dimensions of indicators for parallel analysis, avoiding logical conflicts caused by directly merging indicators with different dimensions. This establishes a standardized difference indicator system under multi-dimensional state monitoring.

[0084] The difference index extraction submodule extracts the temperature difference ratio and humidity difference ratio records of each node pair in the current cycle based on the node difference ratio sequence. It extracts the maximum value and binds it with the node number to form a record structure, and stores it in the index sequence set to establish the state difference index sequence.

[0085] Based on the node difference ratio sequence, the temperature difference ratio field and humidity difference ratio field under each node number are aggregated sequentially. After sorting by period number, extreme value judgment operation is performed on the field values, and the ratio item with the largest value in each group is selected as the representative state offset index. For example, the periodic sequence temperature difference ratio of node pair N01-N02 is [0.15, 0.21, 0.1965], with a maximum temperature difference ratio of 0.21, and the humidity difference ratio sequence is [0.3, 0.5, 0.48], with a maximum humidity difference ratio of 0.5. The corresponding period identifier is recorded as the time index point where the maximum offset occurs. A six-field structure {node number pair, period number, maximum temperature difference ratio, maximum humidity difference ratio, temperature difference sequence, humidity difference sequence} is constructed. This structure is used as the input item for the subsequent fluctuation node screening module. All results are summarized by node number pair and entered into a unified data pool to establish a state difference index sequence.

[0086] Please see Figure 4 The multiple offset determination module includes:

[0087] The threshold comparison submodule extracts the temperature difference ratio and humidity difference ratio for the current period from the node difference ratio sequence based on the temperature difference ratio and humidity difference ratio in the inter-node state difference index sequence, using the following formula:

[0088] ;

[0089] The calculation obtains the offset exceeding the threshold intensity value. Based on whether the intensity value is greater than zero, node pairs whose offset exceeds the trend are filtered out, generating a set of offset exceeding the limit node numbers. Represents node pairs within a period The temperature difference ratio, This indicates its average temperature difference ratio over historical cycles. , These represent the wet difference ratio within the current cycle and its average value under historical cycles, respectively. This is the offset of the intensity value exceeding the threshold. The response threshold multiple is set to 1.8;

[0090] Based on the temperature difference ratio and humidity difference ratio of each node pair within the cycle, the temperature difference ratio and humidity difference ratio of all node pairs in the current cycle are first extracted. Then, the historical average ratio obtained by the sliding window is used as the comparison benchmark, with a preset threshold of 1.8 as the benchmark value for judging the multiple deviation in building environmental monitoring. Subsequently, the ratio difference is calculated for each group of node pairs. The multiple obtained by dividing the current cycle temperature difference ratio by its corresponding historical average is calculated, and the multiple obtained by dividing the current cycle humidity difference ratio by the historical average humidity difference ratio is calculated. The multiple is calculated by summing the two factors and taking their absolute value. Then, a threshold value of 1.8 is subtracted from the result. If the difference is greater than 0, the node pair is considered to exceed the allowable fluctuation threshold and is designated as an offset over-limit node pair. Finally, all node pairs meeting this condition are selected and stored in the offset node set. In actual calculations, if the current cycle temperature difference ratio of node pair N01-N02 is 0.65, corresponding to a historical average temperature difference of 0.42, and the humidity difference ratio is 0.88, with a historical average humidity difference of 0.49, then its temperature difference multiple is... The moisture difference multiple is The sum of the two is Take the absolute value and subtract the threshold. Since the result is greater than 0, this node pair is marked as an offset node. Similar operations are performed on other node pairs. The execution data is shown below:

[0091] Table 3 Example Data Table for Multiple Offset Judgment

[0092]

[0093] As shown in Table 3, the result obtained after calculation using the formula is:

[0094] Node pairs N01-N02: →Exceeded limits;

[0095] Node pairs N02-N03: →Exceeded limits;

[0096] Node pairs N03-N04: →Exceeded limits;

[0097] The above judgment indicates that all node pairs are considered to have fluctuating offsets, and the set of node numbers with offsets exceeding the limit is obtained.

[0098] The offset exceeding threshold intensity value is a quantitative indicator used to measure whether the overall state offset of a node pair in the current detection period exceeds a preset judgment threshold. Specifically, it is calculated by normalizing and comparing the temperature difference ratio and humidity difference ratio of the current period with their historical period averages, calculating the sum of the multiples of the two, taking the absolute value, and then subtracting the set threshold multiple. The result is the offset exceeding threshold intensity value. This value reflects the overall degree to which the state change in the current period exceeds the historical level in terms of temperature and humidity. If the value is greater than zero, it means that the environmental state offset of the node pair in the current period has exceeded the acceptable normal fluctuation range and belongs to the category of abnormal fluctuation. Therefore, this intensity value is not only used to determine whether the offset exceeds the limit, but also to distinguish the offset levels of different node pairs, so as to construct a spatial distribution map and dynamic evolution sequence of the offset degree.

[0099] The formula's operational logic aims to identify whether node pairs have undergone abnormal changes within a cycle by comparing the intensity of the current cycle's state shift with its historical stable states for a standardized judgment. Specifically, the formula first considers the temperature difference ratio of the current cycle... Divide by its average over three historical periods And similarly, the wet difference ratio Divide by its corresponding historical average The two ratios represent the magnitude of change in temperature and humidity compared to historical levels during the current cycle. The two dimensionless ratios are then added together to comprehensively consider the overall deviation in both temperature and humidity dimensions. Using absolute values ​​masks positive and negative directions, uniformly measuring the overall deviation intensity, and finally, this is compared to a preset threshold multiple. (1.8 here) Perform difference calculation to determine whether the fluctuation exceeds the set fluctuation limit. If the final result is positive, it is determined that the node has a multiple offset in the current period.

[0100] The region mapping submodule divides the regions according to the number prefix of the node number pair based on the offset over-limit node number set, extracts the region identifier carried in the number as the region mapping index key, and clusters all over-limit node pairs according to the region number to obtain the region offset node index structure.

[0101] Based on the set of offset node numbers, the prefix number field of each node pair is extracted to identify its monitoring area. The first two digits of the node number name represent the area code. If the node pair number is N01-N02, the area number is R01; if the number is N02-N03, it is R02. All node pairs are numbered and grouped according to this rule. When encountering cross-area node pairs such as N01-N04, their mapping records in the two areas R01 and R04 are recorded respectively. Then, the node pair set is aggregated according to the area number to generate a one-to-many mapping data structure from area to node pair. A mapping table is generated with the area number as the primary key index. After completing the area binding operation, the offset node pair index information under all areas is recorded. For example, node pairs N01-N02 and N01-N03 are mapped to area R01 respectively, and node pairs N03-N04 are mapped to area R03. Finally, the area offset node index structure is obtained.

[0102] The node graph generation submodule obtains the set of node number pairs under each region based on the region offset node index structure, generates node connection edges by indexing the region number, constructs an undirected graph structure according to the order of node pairs, registers the graph structure by indexing the period, and generates a distribution map of node offset multiples within the period.

[0103] The process involves calling the region offset node index structure. First, all offset node pairs contained in each region number are extracted. The starting and target node numbers in each pair are then parsed and used as the two endpoints in the graph, establishing undirected edge connections. During execution, an edge set structure is built region by region, with edges representing connections between nodes that involve offset behavior. Next, a blank graph structure object is created, and the current period number is used as the graph structure's time index key. The extracted node edge sets are then grouped into the graph structure according to region to form an undirected graph structure. This process is then completed. If the region number is R01, and it contains offset node pairs N01-N02 and N01-N03, then the graph structure should contain three nodes N01, N02, and N03, and two edges N01-N02 and N01-N03. Further, if offset node pairs N03-N04 and N04-N05 are detected in region R03, then edges N03-N04 and N04-N05 are formed in the graph, with nodes N03, N04, and N05. This constitutes the offset connection relationship graph under this region. Example data is as follows:

[0104] Table 4. Regional Graph Node Boundary Set Data Table

[0105]

[0106] As shown in Table 4, the offset node map formed by region R01 in the current period contains three nodes and two edges, and the map formed by region R03 contains three nodes and two edges. All nodes and edges are archived with the current period number. Recording the above structure into the map sequence completes the construction of the map under the current period, and finally the multiple offset node distribution map within the period is obtained.

[0107] Please see Figure 5 The fluctuation priority labeling module includes:

[0108] The periodic data extraction submodule extracts the raw temperature and humidity data uploaded by each node in the current period and the previous two periods based on the node number in the node distribution map of multiple offset nodes within the period. The data is then categorized and organized according to the node number, and a temperature and humidity data sequence indexed by the period is generated according to the node number to establish a three-period temperature and humidity data structure.

[0109] Based on the node numbers in the distribution map of node offsets within a period, the index extraction operation is first performed on all node numbers involved in the current period. Then, the original temperature and humidity records uploaded by each node in the current period, the previous period, and the two periods before that are accessed. A data matching process is performed to confirm the data's time of origin, with the time distinguished according to the period number. At the same time, it is ensured that the node number has unique identification capability within the three periods. Then, the temperature and humidity records of each node are periodically mapped by using the period number as the index label. The records are sorted by the upload time within the period. The record with the most stable sampling frequency in each period is selected as the representative value, and missing items or outliers are removed. For example, a node number N01 has 12 records uploaded in the current period, collected at different hours. It also uploaded 10 records in the previous period and 9 records in the two periods before that. After filtering out outliers, 10, 9, and 8 valid records are retained for each period. Then, the average temperature and average humidity of each record in the period are calculated to form a periodic time series data matrix of temperature and humidity. Finally, the mean data of the three periods are integrated with the node number as the primary key to generate the following structure:

[0110] Table 5. Average Temperature and Humidity Data for Three Periods

[0111]

[0112] As shown in Table 5, for node N01, the average temperature for the three periods is 25.4℃, 24.8℃, and 25.0℃, and the average humidity is 56.2%, 52.8%, and 53.5%. Each row in the table is a record of temperature and humidity data for three periods for a node number. The temperature and humidity values ​​are obtained from the raw data uploaded by the node, and are obtained after periodic grouping and record cleaning, resulting in the final three-period temperature and humidity data structure.

[0113] The threshold judgment submodule reads the temperature and humidity data corresponding to the current cycle in terms of nodes based on the three-cycle temperature and humidity data structure, compares it with the same type of data in the previous two cycles, performs difference level confirmation operation for each type of data, sets the standard value for judging temperature and humidity fluctuations, determines whether it exceeds the perceptible range, identifies and archives the node priority status, and obtains the node fluctuation status label set.

[0114] Based on the periodic temperature and humidity data structure, the difference between the current periodic temperature value of each node and the temperature values ​​of the previous and two previous periods is calculated to identify its periodic variation trend in the temperature dimension. The processing flow for the humidity dimension is the same. The judgment criteria are based on the temperature and humidity change perception thresholds proposed in the thermal environment measurement standard, namely, the temperature threshold is ±0.5℃ and the humidity threshold is ±3%RH. If the current periodic value of either dimension exceeds the deviation range between the threshold and the historical periodic average, the node is determined to be a node with a significant fluctuation trend. The judgment process does not involve cumulative or weighted processing, but only relies on the simple comparison difference between the two periodic averages. For example, node N01 in Table 5 has a current periodic temperature... The temperature is 25.4℃, the average of the previous two cycles is (24.8+25.0) / 2=24.9℃, the temperature deviation is 0.5℃, which is on the edge of the perceptible critical point. The humidity in the current cycle is 56.2%, the average of the previous two cycles is (52.8+53.5) / 2=53.15%, the humidity deviation is 3.05%, which exceeds the set threshold. Therefore, node N01 is determined to be a fluctuating node. If the current cycle temperature of node N03 is 24.7℃, the historical average is 24.85℃, the deviation is 0.15℃, and the humidity deviation is also within 1%, then N03 is a stable node. After all nodes have been compared, status labels are assigned according to the node number, and finally a set of node fluctuation status labels is generated.

[0115] The priority identifier generation submodule extracts the node number and status identifier content marked as fluctuating state in the current cycle based on the node fluctuation state label set, and establishes a time index in combination with the cycle number. It then outputs the status result information of each node in a structured manner, maps and integrates the node number and priority identifier, and establishes a node-level temperature and humidity fluctuation priority identifier set.

[0116] For all node numbers included in the node fluctuation status label set, an identifier structure is constructed and indexed according to the period number. The node numbers that are determined to have temperature and humidity fluctuation status in the current period are extracted. After binding the node with its corresponding label information, the nodes are archived. The labels are divided into two categories: high priority and ordinary priority. No specific numerical calculation is involved. The labels are directly inherited according to the aforementioned fluctuation judgment criteria. A period-node number-priority label triple structure is generated and the node identifier set is stored in a hash structure for subsequent calls. If nodes N01 and N02 are marked as high priority during the structure construction process, they are recorded in the set as N01: high priority and N02: high priority. Conversely, if N03 is marked as ordinary priority, it is recorded as N03: ordinary priority. After sorting, a node-level temperature and humidity fluctuation priority identifier set is established.

[0117] Please see Figure 6 The data processing and scheduling module includes:

[0118] The priority pool partitioning submodule is based on the priority tag information of all nodes in the node-level temperature and humidity fluctuation priority identifier set. It writes the data unit corresponding to the high-priority node number into the priority scheduling processing pool and writes the data unit corresponding to the ordinary priority node number into the ordinary scheduling processing pool. It sorts the data in ascending order according to the timestamp carried by the node number, removes data entries with time conflicts, and caches the data sets of each priority type in a queue structure to generate a priority scheduling dataset.

[0119] Based on the high-priority and ordinary-priority label fields already categorized in the node-level temperature and humidity fluctuation priority identifier set, all node numbers are first extracted, and the nodes are divided into two groups according to their label fields. Then, the data units corresponding to each node number are obtained. Each data unit must contain a period number, timestamp, temperature and humidity value, and label identifier field. After acquisition, high-priority data units are written to a high-priority queue cache structure, which must meet the system setting of a response latency not exceeding 100 milliseconds. Ordinary-priority data units are written to a regular scheduling queue, meeting a response limit of 500 milliseconds. The write operation is based on the time of each data unit. The timestamp field is sorted in ascending order, and invalid or duplicate records are removed. During this process, data with an upload time interval of less than 10 milliseconds is considered a duplicate record and merged. For example, if the upload record timestamps for node N01 in the two priority queues are 14:00:01.004 and 14:00:01.006 respectively, they are considered records within the same period, and the former is selected for retention. After processing all data in the above manner, two complete scheduling pools are formed. In addition to the timestamp, the node number and data content are also recorded in the scheduling pool structure, and a priority label field is attached for reference in subsequent processing. The final structure is as follows:

[0120] Table 6. Priority Scheduling Pool Example Structure Table

[0121]

[0122] As shown in Table 6, priority labels and time order have formed a clear classification. Based on this, they are uniformly encapsulated into two sets of structures: priority scheduling processing pool and ordinary scheduling processing pool, to obtain the priority scheduling dataset.

[0123] The time queue management submodule is based on the priority scheduling dataset. It sets up two scheduling processing channels according to the scheduling priority type. In each scheduling cycle, it performs dequeue processing on the data unit at the head of the queue. Each processing records the current system timestamp and uses the time as the scheduling time point of the data unit. It records the dequeue sequence number and organizes the records with the node number as the primary key to obtain the data processing scheduling sequence record table.

[0124] Data from each node in the priority scheduling dataset is retrieved and entered into two scheduling channels: a high-priority queue and a normal-priority queue. The high-priority queue has a processing cycle of 20 milliseconds, while the normal-priority queue has a processing cycle of 100 milliseconds. All data units are dequeued in ascending order of their timestamps, and the dequeue order becomes the task execution order number. The processing cycle unit is set according to the system clock frequency configured on the task execution platform. At the start of scheduling, each dequeue operation records the current system time as the scheduling time point and also marks the processing resource type used by the task. For example, task N01's dequeue time in the high-priority queue is 14:00:01.020. The resource type is edge node A1, and the scheduling sequence number is 1. The scheduling record table needs to bind the node number N01, the time point 14:00:01.020, the resource type A1, and the scheduling sequence number 1. After processing, the record structure is immediately written into the memory structure for caching. All records are dynamically added according to the scheduling time order. The structure fields include task number, time point, resource type, and scheduling sequence number. There is no data calculation. Only the fields are archived and bound in order. Priority judgment is no longer performed in each round of dequeueing. Only the sequential flow processing process is executed. All records are finally collected and a data processing scheduling sequence record table is generated.

[0125] The scheduling record generation submodule, based on the node number, scheduling time point, processing resource type, and scheduling sequence number of each record in the data processing scheduling sequence record table, arranges all fields in ascending order according to the node number, and uses the cycle number and processing time point to generate a unique primary key index to distinguish the task scheduling behavior at different time points in the same cycle, and establishes a distributed indoor temperature and humidity data processing record table.

[0126] Based on the task field data already written in the data processing scheduling sequence record table, all node numbers are used as the primary key of the index. The corresponding scheduling time point, processing resource type, and scheduling sequence number fields are extracted one by one and uniformly written into a standard record structure template. The structure template uses the cycle number and scheduling time to generate a unique index identifier. An example of this field combination is "20240809_140001020". This structure is used to prevent duplicate writing of scheduling records within the same time period. After field unification, a bidirectional index mapping structure is established between the node number field and the resource type field. That is, one node number can correspond to multiple processing resource type records, and multiple resource types can also be used to locate their corresponding node numbers. Finally, the above information is written into a table structure and arranged in order of scheduling time point. A uniform output structure example is as follows:

[0127] Table 7. Sample Distributed Processing Records

[0128]

[0129] As shown in Table 7, the node number and its scheduling behavior record have been structurally encapsulated and uniformly written into the system's main record file. The record content has periodic index and scheduling sequence number information, and finally a distributed indoor temperature and humidity data processing record table is established.

[0130] A distributed indoor temperature and humidity data processing method includes the following steps:

[0131] S1: Obtain the node number of the indoor area, collect the temperature and humidity sampling values ​​of the current period respectively, establish a one-to-one corresponding data set according to the node number, bind the node number pair with the corresponding temperature and humidity data, and generate the node difference initialization dataset.

[0132] S2: Initialize the dataset based on node differences, calculate the difference between the temperature and humidity sampled values ​​of each pair of nodes, bind them to the node number pair in sequence to form a difference record unit, perform difference operation between the difference record unit and the average difference of the node's historical three periods, extract the maximum value of the temperature difference ratio and humidity difference ratio, and obtain the state difference index sequence between nodes.

[0133] S3: Based on the sequence of state difference indicators between nodes, select all node pairs whose maximum ratio value exceeds the response threshold multiple and number them, and establish a corresponding regional number mapping to obtain the distribution map of nodes with multiple offset within the period.

[0134] S4: Based on the distribution map of nodes with multiple offsets within the cycle, calculate the average temperature and humidity data of each node in the current cycle and the previous two cycles. If it exceeds the fluctuation sensitivity threshold, mark it as a high-priority data unit; otherwise, mark it as a normal-priority unit and generate a set of node-level temperature and humidity fluctuation priority identifiers.

[0135] S5: Based on the node-level temperature and humidity fluctuation priority identifier set, insert them into the priority scheduling processing pool and the normal scheduling processing pool according to priority type, and dequeue them in time order to execute data processing sequentially, and establish a distributed indoor temperature and humidity data processing record table.

[0136] The above are merely preferred embodiments of the present invention and are not intended to limit the present invention in any other way. Any person skilled in the art may make changes or modifications to the above-disclosed technical content to create equivalent embodiments that can be applied to other fields. However, any simple modifications, equivalent changes, and modifications made to the above embodiments based on the technical essence of the present invention without departing from the scope of the present invention shall still fall within the protection scope of the present invention.

Claims

1. A distributed indoor temperature and humidity data processing system, characterized in that, The system includes: The node data initialization module obtains the node number of the indoor area, collects the temperature and humidity sampling values ​​of the current period, establishes a one-to-one corresponding data set based on the node number, and binds the node number pair with the corresponding temperature and humidity data to generate the node difference initialization dataset. The state difference extraction module initializes the dataset based on the node difference, calculates the difference between the temperature and humidity sampling values ​​of each pair of nodes, binds them to the node number pairs in sequence to form difference record units, performs difference calculation between the difference record units and the average difference of the node's historical three periods, extracts the maximum value of the temperature difference ratio and humidity difference ratio, and obtains the state difference index sequence between nodes. The multiple offset determination module filters out all node pairs whose maximum ratio value exceeds the response threshold multiple based on the state difference index sequence between nodes, and establishes a corresponding regional number mapping to obtain the distribution map of multiple offset nodes within the period. The fluctuation priority labeling module calculates the average temperature and humidity data of each node in the current cycle and the previous two cycles based on the distribution map of the multiple offset nodes within the cycle. If the average temperature and humidity data exceeds the fluctuation sensitivity threshold, the node is marked as a high-priority data unit; otherwise, it is marked as a normal-priority unit, thus generating a set of node-level temperature and humidity fluctuation priority identifiers. The node data initialization module includes: The node number extraction submodule obtains all adjacent node numbers from the indoor area node number table, identifies each pair of node number pairs with a direct connection relationship, and establishes a set of number pairs to generate a node number pair set. The environmental data acquisition submodule reads the temperature and humidity sampling values ​​corresponding to each pair of node numbers in the current period based on the set of node number pairs, extracts the environmental sampling values ​​of each node, and combines them to construct a double set of environmental datasets under the node pair, thus obtaining a double set of environmental sampling values. The difference data generation submodule extracts the sequence number, period number and environmental data combination information of each pair of nodes based on the two sets of environmental sampling values, organizes them in a unified structure, binds each node number pair to its corresponding environmental data set, and generates the node difference initialization dataset. The multiple offset determination module includes: The threshold comparison submodule extracts the temperature difference ratio and humidity difference ratio of the current period from the node difference ratio sequence based on the temperature difference ratio and humidity difference ratio in the inter-node state difference index sequence, calculates and obtains the offset exceeding the threshold intensity value, and filters out node pairs whose offset exceeds the trend based on whether the intensity value is greater than zero, and generates a set of offset exceeding the limit node numbers. The region mapping submodule divides the regions according to the number prefix of the node number pair based on the offset over-limit node number set, extracts the region identifier carried in the number as the region mapping index key, and clusters all over-limit node pairs according to the region number to obtain the region offset node index structure. The node graph generation submodule obtains the set of node number pairs under each region according to the region offset node index structure, generates node connection edges according to the region number as the index, constructs an undirected graph structure according to the order of node pairs, registers the graph structure with the period as the index number, and generates a node distribution map of multiple offset within the period.

2. The distributed indoor temperature and humidity data processing system according to claim 1, characterized in that, The node difference initialization dataset includes a node number pair mapping relationship, a temperature and humidity sampling dataset within the period, and a data set binding structure. The state difference index sequence includes a temperature difference ratio sequence between nodes, a humidity difference ratio sequence between nodes, and a record of the maximum value of the ratio difference. The node distribution map of multiple offset within the period includes node pair numbers exceeding the threshold, a node region number mapping relationship, and an abnormal distribution map structure within the period. The node-level temperature and humidity fluctuation priority identifier set includes the node temperature and humidity mean, the comparison result with the fluctuation sensitivity threshold, and priority identifier labels.

3. The distributed indoor temperature and humidity data processing system according to claim 1, characterized in that, The state difference extraction module includes: The difference record generation submodule initializes the dataset based on the node difference, extracts the current cycle temperature and humidity sampling values ​​corresponding to each pair of node numbers, establishes a set of difference data for the temperature difference and humidity difference values ​​of each pair of nodes, unifies the format of all records and includes them in the cache set, and establishes node environment difference records. The periodic difference comparison submodule, based on the node environmental difference record, uses the node number pair and the current period number as index keys to read the historical data of temperature difference and humidity difference values ​​of the node number pair in the previous three consecutive periods. It then processes the difference ratio between the current period temperature difference value and the average temperature difference value of the previous three periods, and performs the same operation on the current period humidity difference value and the average humidity difference value of the previous three periods. The periodic temperature difference ratio and periodic humidity ratio are calculated and obtained respectively, and then uniformly bound to the node number pair and the period number to obtain the node difference ratio sequence. The difference index extraction submodule extracts the temperature difference ratio and humidity difference ratio records of each node pair in the current cycle based on the node difference ratio sequence, extracts the maximum value and binds it with the node number to form a record structure, and stores it in the index sequence set to establish a state difference index sequence.

4. The distributed indoor temperature and humidity data processing system according to claim 1, characterized in that, The fluctuation priority labeling module includes: The periodic data extraction submodule extracts the raw temperature and humidity data uploaded by each node in the current period and the previous two periods based on the node number in the distribution map of the multiple offset nodes within the period. The data is then categorized and organized according to the node number, and a temperature and humidity data sequence indexed by the period is generated according to the node number to establish a three-period temperature and humidity data structure. The threshold judgment submodule reads the temperature and humidity data corresponding to the current cycle in terms of nodes according to the three-cycle temperature and humidity data structure, compares it with the same type of data in the previous two cycles, performs difference level confirmation operation for each type of data, sets temperature and humidity fluctuation judgment standard values, determines whether it exceeds the perceptible range, identifies and archives the node priority status, and obtains the node fluctuation status label set. The priority identifier generation submodule extracts the node number and status identifier content marked as fluctuating in the current cycle based on the node fluctuation status label set, and establishes a time index in combination with the cycle number. It then outputs the status result information of each node in a structured manner, maps and integrates the node number and priority identifier, and establishes a node-level temperature and humidity fluctuation priority identifier set.

5. The distributed indoor temperature and humidity data processing system according to claim 1, characterized in that, The system also includes: The data processing and scheduling module inserts data into the priority scheduling processing pool and the normal scheduling processing pool according to the priority type based on the node-level temperature and humidity fluctuation priority identifier set, and executes the data processing sequentially according to the time order, thereby establishing a distributed indoor temperature and humidity data processing record table. The distributed indoor temperature and humidity data processing record table includes node number record items, scheduling time point information, processing resource type identifier, and task execution order record items.

6. The distributed indoor temperature and humidity data processing system according to claim 5, characterized in that, The data processing scheduling module includes: The priority pool partitioning submodule, based on the priority tag information of all nodes in the node-level temperature and humidity fluctuation priority identifier set, writes the data unit corresponding to the high-priority node number into the priority scheduling processing pool and writes the data unit corresponding to the ordinary priority node number into the ordinary scheduling processing pool. It sorts the data in ascending order according to the timestamp carried by the node number, removes data entries with time conflicts, and caches the data sets of each priority type in a queue structure to generate a priority scheduling dataset. The time queue management submodule sets up two scheduling processing channels based on the priority scheduling dataset and the scheduling priority type. In each scheduling cycle, it performs dequeue processing on the data unit at the head of the queue. Each processing records the current system timestamp and uses the time as the scheduling time point of the data unit. It records the dequeue sequence number and organizes the records with the node number as the primary key to obtain the data processing scheduling sequence record table. The scheduling record generation submodule, based on the node number, scheduling time point, processing resource type, and scheduling sequence number information of each record in the data processing scheduling sequence record table, arranges all fields in ascending order according to the node number, and uses the cycle number and processing time point to generate a unique primary key index to distinguish the task scheduling behavior at different time points in the same cycle, and establishes a distributed indoor temperature and humidity data processing record table.

7. A distributed indoor temperature and humidity data processing method, characterized in that, The method is used to implement the distributed indoor temperature and humidity data processing system according to any one of claims 1-6, and includes the following steps: S1: Obtain the node number of the indoor area, collect the temperature and humidity sampling values ​​of the current period respectively, establish a one-to-one corresponding data set according to the node number, bind the node number pair with the corresponding temperature and humidity data, and generate the node difference initialization dataset. S2: Based on the node difference initialization dataset, calculate the difference of temperature and humidity sampling values ​​for each pair of nodes, bind them sequentially with node number pairs to form difference record units, perform difference operation between the difference record units and the average difference of the node's historical three periods, extract the maximum value of temperature difference ratio and humidity difference ratio, and obtain the state difference index sequence between nodes. S3: Based on the state difference index sequence between nodes, filter all node pairs whose maximum ratio value exceeds the response threshold multiple, and establish a corresponding regional number mapping to obtain the distribution map of nodes with multiple offset within the period. S4: Based on the distribution map of the multiple offset nodes within the cycle, calculate the average temperature and humidity data of each node in the current cycle and the previous two cycles. If it exceeds the fluctuation sensitivity threshold, mark it as a high-priority data unit; otherwise, mark it as a normal-priority unit and generate a set of node-level temperature and humidity fluctuation priority identifiers. S5: Based on the node-level temperature and humidity fluctuation priority identifier set, insert them into the priority scheduling processing pool and the normal scheduling processing pool according to priority type, and dequeue them in time order to perform data processing in sequence, and establish a distributed indoor temperature and humidity data processing record table.

Images