Architectural decoration intelligent environment monitoring system based on Internet of Things

Abstract

The invention provides a building decoration intelligent environment monitoring system based on the Internet of Things, and belongs to the technical field of the Internet of Things, and the system comprises a sensor layer which is disposed in a building decoration area and is used for collecting environment parameters in a building decoration environment in real time, and an Internet of Things communication layer which is disposed in the building decoration area. The data transmission layer is used for transmitting data acquired by the sensor layer to the data processing layer according to a transmission standard of the sensor layer; the data processing layer is used for preprocessing real-time data, analyzing and determining a highlighted state change set of each environmental parameter based on a data curve based on the preprocessed data, and modeling and establishing a building decoration environment quality model; and the data processing layer is used for analyzing the building decoration environment quality model according to a preset environment quality standard to generate an environment monitoring report and early warning information, and issuing the environment monitoring report and the early warning information to the application layer to control execution equipment to operate. The problems of uneven coverage, low transmission efficiency, one-sided evaluation and the like in traditional monitoring are solved, and the accuracy and the intelligent level of building decoration environment monitoring are remarkably improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of Internet of Things, and particularly relates to a building decoration intelligent environment monitoring system based on Internet of Things. BACKGROUND

[0002] In the field of building decoration, the indoor environment quality directly affects the health and comfort of the occupants, and the real-time monitoring of environmental parameters such as temperature and humidity, formaldehyde concentration, PM2.5 concentration, etc. is the core link to ensure environmental quality. Traditional building decoration environment monitoring relies on manual inspection or single-point fixed sensors, which has the following significant limitations: traditional monitoring methods use single parameter sensors deployed in a scattered manner, lacking systematic layout planning, resulting in problems such as data collection redundancy or missing coverage. For example, the release patterns of volatile pollutants such as formaldehyde concentration in different decoration material areas differ significantly, but fixed-point sensors are difficult to capture spatial heterogeneity characteristics, and manual inspection has low frequency and poor timeliness, which cannot reflect the dynamic changes of environmental parameters in real time. Moreover, the existing transmission method is real-time collection and real-time transmission, which will invisibly waste communication resources, resulting in low data interaction efficiency. In addition, current pre-alarm is triggered by single parameter exceeding, which is prone to misjudgment or omission, reducing the accuracy of environmental quality assessment. In summary, various factors make it difficult to meet the needs of precise and intelligent monitoring of building decoration environment.

[0003] Therefore, the present application provides a building decoration intelligent environment monitoring system based on Internet of Things. SUMMARY

[0004] The present application provides a building decoration intelligent environment monitoring system based on Internet of Things to solve the above technical problems.

[0005] The present application provides a building decoration intelligent environment monitoring system based on Internet of Things, comprising:

[0006] A sensor layer deployed in the building decoration area for real-time collection of environmental parameters in the building decoration environment, including temperature and humidity, formaldehyde concentration, PM2.5 concentration, illuminance and carbon dioxide concentration;

[0007] An Internet of Things communication layer for transmitting data collected by the sensor layer to the data processing layer according to the transmission standard of the sensor layer;

[0008] A data processing layer for preprocessing real-time data collected by the sensor, including data filtering, outlier detection and data normalization processing, analyzing and determining the prominent state change set of each environmental parameter based on the data curve based on the preprocessed data, and modeling to establish a building decoration environment quality model;

[0009] The data processing layer is further configured to analyze the building decoration environment quality model according to a preset environment quality standard, generate an environment monitoring report and early warning information, and send the environment monitoring report and early warning information to the application layer to control the execution equipment to operate and improve the environment quality.

[0010] Preferably, the method further comprises: a cutting module configured to cut the three-dimensional region structure of the building decoration environment according to region functions to obtain a plurality of cutting spaces, and to determine a refinement unit for further cutting according to building distribution and decoration distribution of each cutting space;

[0011] a set construction module configured to generate a static environment set of each further cutting unit according to a first derivative environment factor of a building type involved in the further cutting unit and a second derivative environment factor of a decoration type involved in the further cutting unit in the same cutting space;

[0012] a rough determination module configured to analyze environment monitoring importance of each static environment set, and to lock a rough installation height range and a rough installation horizontal range according to the environment monitoring importance;

[0013] a single installation range determination module configured to construct a factor distribution structure of each derivative environment factor in the corresponding cutting space based on all static environment sets in the same cutting space, and to analyze each layer factor dense trend and each column factor dense trend in the factor distribution structure to lock a to-be-installed height range and a to-be-installed horizontal range corresponding to the corresponding derivative environment factor;

[0014] a negotiation processing module configured to perform range negotiation processing on all to-be-installed height ranges and to-be-installed horizontal ranges and the rough installation height range and the rough installation horizontal range respectively to determine a first range set with the most negotiation times in the rough installation height range and a second range set with the most negotiation times in the rough installation horizontal range, wherein the range set is empty or contains at least one negotiation range segment, and when the range set is empty, a range overlap segment with the highest number of times corresponding to all to-be-installed ranges is determined;

[0015] a distribution module configured to determine a first concentrated distribution of a sensor group for a key space according to a first final segment in a vertical direction and a first ratio of a vertical height, a second final segment in a horizontal direction and a second ratio of a horizontal length, and in combination with a relative direction of the first final segment and the second final segment, and to perform a second uniform distribution of the sensor group in a non-key space, wherein the key space is determined based on the first final segment and the second final segment, and the non-key space is a remaining space in the corresponding cutting space except the key space;

[0016] The distribution result of the sensor group is used as a sensor layer deployed in the building decoration region.

[0017] Preferably, the cutting module comprises:

[0018] a coefficient determination unit configured to determine a cutting coefficient of a refinement unit according to an architectural distribution and a decoration distribution of each cutting space;

[0019]

[0020] wherein U is the cutting coefficient of the refinement unit; and a is a weight coefficient of the architectural distribution of the corresponding cutting space; represents an architectural distribution complexity index, Vii is the volume of an ith1 architectural component in the corresponding cutting space, Vo is the total volume of the corresponding cutting space, gii is an environmental interference coefficient of the ith1 architectural component in the corresponding cutting space, and n1 is the total number of architectural components in the corresponding cutting space; and b is a weight coefficient of the decoration distribution of the corresponding cutting space; is a decoration distribution sensitivity index, Si2 is the surface area of an ith2 decoration component in the corresponding cutting space, So is the total surface area of the corresponding cutting space, fi2 is an environmental release coefficient of the ith2 decoration component in the corresponding cutting space; Fs is a space function correlation factor; Gk is a space connectivity interference coefficient; g is a space topology weight coefficient of the corresponding cutting space; S1 is the effective ventilation area of a door and a window of the corresponding cutting space; S2 is the total wall area of the corresponding cutting space; and fq is an air flow disturbance coefficient of the corresponding cutting space;

[0021] a re-cutting unit configured to re-cut the corresponding cutting space by rounding up the cutting coefficient.

[0022] Preferably, the system further comprises:

[0023] a starting module configured to start the sensor layer to collect multi-modal collection of the architectural decoration area and synchronously collect operation characteristic data of each sensor group;

[0024] an image analysis module configured to collect an initial thermal imaging image of each sensor group by the infrared sensor deployed in the architectural decoration area, and adjust the initial thermal imaging image to obtain a real-time thermal imaging image in combination with temperature and humidity data;

[0025] a basic analysis module configured to generate an analysis pair based on the operation characteristic data of each sensor group and the real-time thermal imaging image, periodically divide a series of analysis pairs collected in real time, and generate a basic feature set;

[0026] a difference analysis module configured to predict a temperature change curve of the sensor group according to the operation characteristic data of the sensor group, and perform time alignment processing with the real-time thermal imaging image to obtain a temperature difference feature set;

[0027] A threshold setting module is configured to generate a transmission trigger threshold based on the basic feature set and the temperature difference feature set, and to divide all sensor groups into different levels according to the transmission trigger threshold. The sensor groups after the division are synchronously triggered according to the minimum trigger threshold of the corresponding level, and the collected data is transmitted from the sensor layer to the data processing layer based on the Internet of Things communication layer.

[0028] Preferably, the threshold setting module comprises:

[0029] A pulse acquisition unit is configured to acquire the historical interruption duration of each sensor group during the working process, and to acquire each two continuous measurement pulses before and after the historical interruption duration. The time width of the first two continuous measurement pulses is compared with that of the last two continuous measurement pulses to obtain a basic error correction width.

[0030] A clock deviation determination unit is configured to synchronously acquire the clock deviation data of each sensor device of each sensor group, to calculate the frequency compensation of the clock of each sensor device relative to the reference clock of the sensor group, and to construct a clock deviation sequence of the sensor group.

[0031] A symbol determination unit is configured to obtain an initial trigger threshold by associating and comparing the basic feature set and the temperature difference feature set, and to assign a judgment symbol to the basic error correction width according to the direction of the basic feature set and the temperature difference feature set, and in combination with the clock deviation sequence, the maximum deviation and the minimum deviation in the clock deviation sequence.

[0032] A threshold acquisition unit is configured to obtain a transmission trigger threshold according to the initial trigger threshold, the judgment symbol and the error correction width.

[0033] Preferably, the data processing layer comprises:

[0034] An initial set determination module is configured to perform single-parameter curve feature extraction on the preprocessed environmental parameter data, to capture the slope mutation point, the continuous abnormal interval and the periodic fluctuation amplitude of each curve by using an adaptive sliding window algorithm, and to generate a single-parameter initial change set.

[0035] A mutation set determination module is configured to calculate the dynamic cross-correlation coefficient of any two environmental parameter curves, to screen the parameter pairs with a correlation coefficient exceeding a threshold, to include the cooperative change features of the corresponding curves into the initial change set, and to form a prominent state change set with fused correlation information.

[0036] Preferably, the data processing layer further comprises:

[0037] The segmented analysis module is used for curve segmentation on the pretreated data, divides analysis subspaces according to the functional characteristics and material characteristics of the building decoration area, and extracts significant change nodes of each environmental parameter curve in each subspace through an extreme point clustering algorithm.

[0038] The subspace set determination module is used for, for each subspace, combining the physical attribute parameters and the space structure parameters of the decoration material, establishing a mapping relationship of material-structure-environmental parameter, filtering the interference changes in the subspace through the mapping relationship, retaining the prominent state changes that are strongly related to the decoration characteristics, and forming a subspace exclusive change set.

[0039] The model determination module is used for, based on the subspace exclusive change set and a required change set of the subspace involved extracted from the prominent state change set, constructing a main model and a submodel hierarchical structure to form a dynamically adapted building decoration environmental quality model.

[0040] Preferably, the main model learns the common environmental quality law across spaces, and the submodel performs parameter fine-tuning for the decoration scene.

[0041] Compared with the prior art, the application has the following beneficial effects:

[0042] Through optimization of the sensor layout, accurate environmental parameter acquisition is realized, through the efficient communication layer, data transmission redundancy is reduced, through the intelligent data processing, a multi-dimensional quality evaluation model is constructed, and finally through the application layer, closed-loop regulation and control of the environment is realized, solving the problems of uneven coverage, low transmission efficiency and one-sided evaluation in the traditional monitoring, and significantly improving the accuracy and intelligent level of the building decoration environmental monitoring.

[0043] Other features and advantages of the present application will be set forth in the following description, and in part will become apparent to those skilled in the art from the description, or can be learned by practice of the present application. The objects and other advantages of the present application will be realized and achieved by means of the structures particularly pointed out in the written description and the appended drawings.

[0044] The technical solutions of the present application will be further described in detail below with the help of the drawings and examples. BRIEF DESCRIPTION OF DRAWINGS

[0045] The accompanying drawings are included to provide a further understanding of the present application, and constitute a part of the specification, together with the embodiments of the present application, to explain the present application, and do not constitute a limitation on the present application. In the drawings:

[0046] Figure 1 The structure diagram of the building decoration intelligent environmental monitoring system based on the Internet of Things in the embodiments of the present application. DETAILED DESCRIPTION

[0047] The preferred embodiments of the present application are described below in conjunction with the accompanying drawings, it should be understood that the preferred embodiments described herein are merely intended to illustrate and explain the present application, and are not intended to limit the present application.

[0048] The present application provides a building decoration intelligent environment monitoring system based on Internet of Things, as shown in the figure, comprising: Figure 1

[0049] The sensor layer deployed in the building decoration area is used to collect environmental parameters in the building decoration environment in real time, including temperature and humidity, formaldehyde concentration, PM2.5 concentration, illuminance and carbon dioxide concentration.

[0050] The Internet of Things communication layer is used to transmit the data collected by the sensor layer to the data processing layer according to the transmission standard of the sensor layer.

[0051] The data processing layer is used to preprocess the real-time data collected by the sensor, including data filtering, outlier detection and data normalization processing, analyze and determine the prominent state change set of each environmental parameter based on the data curve based on the preprocessed data, and model the building decoration environment quality model.

[0052] The data processing layer is also used to analyze the building decoration environment quality model according to the pre-set environmental quality standard to generate environmental monitoring report and early warning information, and issue it to the application layer control execution equipment for operation to improve the environmental quality.

[0053] In this embodiment, the sensor layer is a hardware collection composed of a plurality of sensors deployed in the building decoration area, which is the front end of the system sensing the environment, and the sensor types involved in the sensor group are the same, and the layout strategy of dense key area and uniform non-key area is adopted.

[0054] The Internet of Things communication layer is a communication network and protocol set for realizing data transmission between the sensor layer and the data processing layer, and the transmission standard is related to the trigger threshold of the sensor group involved in the sensor layer.

[0055] Preprocessing is a process of cleaning the original data to ensure data reliability, including:

[0056] Data filtering: remove high-frequency noise (such as sensor transient fluctuation).

[0057] Outlier detection: identify and eliminate data deviating from the normal range (such as sensor failure).

[0058] Data normalization: standardize different magnitude parameters (such as mapping to 0-1 range) for subsequent analysis.

[0059] In this embodiment, the prominent state change set is a set of change nodes with significant meaning in the environmental parameter data curve.​

[0060] The architectural decoration environment quality model is a quantitative evaluation model integrating multi-parameter correlation characteristics, reflecting the relationship between environment quality and decoration materials and space functions.

[0061] The preset environment quality standard is an evaluation benchmark developed with reference to national standards, industry specifications and user requirements, for example, GB / T 18883-2002 "Indoor Air Quality Standard" is cited to set the formaldehyde ≤0.1 mg / m 3 , CO2 ≤100 ppm; combined with user requirements, the custom standard of child room illuminance ≥200 lux is added.

[0062] The environment monitoring report is a structured document generated based on model analysis, including environment quality status, problem tracing and optimization suggestions, for example, the report shows that the formaldehyde concentration in the bedroom is 0.12 mg / m 3 (20% over standard), mainly due to the release of new wardrobe, and it is recommended to ventilate for 2 hours a day and place activated carbon.

[0063] The early warning information is a prompt signal when the environment parameter is abnormal or there is potential risk, specifically: set multi-level early warning (level 1: slight deviation, level 2: significant over standard, level 3: emergency risk), notify the user through APP push, sound and light alarm and other ways, for example, when the PM2.5 concentration rises to 150 μg / m 3 , the level 2 early warning is triggered, and the push is: it is recommended to close the door and window and turn on the air purifier.

[0064] In this embodiment, the application layer is a terminal system that receives data processing layer instructions and performs environment control, including execution equipment and user interaction interface.

[0065] The execution equipment is a hardware device for improving environment quality, for example, fresh air system (reducing formaldehyde and CO2 concentration), humidifier (adjusting humidity), curtain motor (controlling illuminance), air purifier (reducing PM2.5 concentration), specifically, the equipment is linked with the system through the Internet of Things module, and automatically runs after receiving the instruction (such as the formaldehyde over standard, the fresh air system automatically switches to "strong wind mode").

[0066] The beneficial effects of the above technical solution are: through optimizing the sensor layout to realize accurate collection of environment parameters, relying on efficient communication layer to reduce data transmission redundancy, with the help of intelligent data processing to build a multi-dimensional quality evaluation model, and finally through the application layer to realize the closed-loop control of the environment, solving the problems of uneven coverage, low transmission efficiency and one-sided evaluation in traditional monitoring, significantly improving the accuracy and intelligent level of architectural decoration environment monitoring.

[0067] The present application provides a kind of based on Internet of Things architectural decoration intelligent environment monitoring system, still include:

[0068] The cutting module is configured to cut the three-dimensional region structure of the building decoration environment according to region functions to obtain a plurality of cutting spaces, and to determine a refined unit for further cutting according to building distribution and decoration distribution of each cutting space.

[0069] The set construction module is configured to generate a static environment set of each further cutting unit according to a first derived environment factor of a building type involved in the further cutting unit and a second derived environment factor of a decoration type involved in the further cutting unit in the same cutting space.

[0070] The rough determination module is configured to analyze environment monitoring importance of each static environment set, and to lock a rough installation height range and a rough installation horizontal range according to the environment monitoring importance.

[0071] The single installation range determination module is configured to construct a factor distribution structure of each derived environment factor in the corresponding cutting space based on all static environment sets in the same cutting space, and to analyze each layer factor dense trend and each column factor dense trend in the factor distribution structure to lock a to-be-installed height range and a to-be-installed horizontal range corresponding to the corresponding derived environment factor.

[0072] The negotiation processing module is configured to perform range negotiation processing on all to-be-installed height ranges and to-be-installed horizontal ranges and the rough installation height range and the rough installation horizontal range respectively, to determine a first range set with the most negotiation times in the rough installation height range and a second range set with the most negotiation times in the rough installation horizontal range, wherein the range set is empty or contains at least one negotiation range segment, and when the range set is empty, a range overlap segment based on the highest number of corresponding all to-be-installed ranges is determined.

[0073] The distribution module is configured to determine a first concentrated distribution of a sensor group for a key space according to a first final segment in a vertical direction and a first ratio of a vertical height, a second final segment in a horizontal direction and a second ratio of a horizontal length, and in combination with a relative direction of the first final segment and the second final segment, and to perform a second uniform distribution of the sensor group in a non-key space, wherein the key space is determined based on the first final segment and the second final segment, and the non-key space is a remaining space in the corresponding cutting space except the key space.

[0074] The distribution result of the sensor group is used as a sensor layer deployed in the building decoration region.

[0075] In this embodiment, the three-dimensional region structure is a three-dimensional structure of the building decoration environment in a three-dimensional space, containing length, width and height information.

[0076] Regional function is the function that a region of a building decoration assumes, for example, the regional function of a living room is to receive visitors and hold family activities; the regional function of a bedroom is to rest and sleep.

[0077] Cutting space is a space unit obtained by cutting a three-dimensional regional structure according to regional functions, for example, a large office space is cut into cutting spaces such as an office area, a meeting area, and a tea room according to regional functions.

[0078] Building distribution is the layout of building components in a cutting space, the building components including walls, beams, columns, floors, etc., for example, in a cutting space, there are two solid walls and one load-bearing column, the positions and arrangements of the walls and the column are the building distribution of the space.

[0079] Decoration distribution is the layout of decoration components in a cutting space, the decoration components including furniture, paint, soft furnishings, etc., for example, in a cutting space, there is a sofa, a TV cabinet, and a wall painted with latex paint, the positions of these decoration components constitute the decoration distribution of the space.

[0080] Refinement unit is a smaller space unit obtained by further cutting a cutting space according to building distribution and decoration distribution, for example, a living room cutting space is further cut according to the decoration distribution of a sofa, a TV wall, and a coffee table, to obtain a sofa peripheral refinement unit and a TV wall area refinement unit.

[0081] Further cutting unit is a space unit obtained by further cutting a cutting space by a cutting module, which is a more detailed division than a cutting space, for example, a kitchen cutting space is further cut into a cooking range peripheral unit and a sink peripheral unit.

[0082] Building type is the type of building component involved in a further cutting unit, for example, building types can include masonry walls, concrete beams and columns, and glass curtain walls.

[0083] First derivative environmental factor is an environmental factor influenced by a building type, for example, the first derivative environmental factors of a masonry wall can include thermal insulation (affecting temperature and humidity), air permeability (affecting air circulation), etc.

[0084] Decoration type is the type of decoration component involved in a further cutting unit, for example, decoration types can include solid wood furniture, cloth curtains, and latex paint walls.

[0085] Second derivative environmental factor is an environmental factor influenced by a decoration type, for example, the second derivative environmental factors of solid wood furniture can include formaldehyde release amount and moisture absorption rate (affecting humidity), etc.; the second derivative environmental factors of latex paint walls can include VOC release amount, etc.

[0086] The static environment set is a set of the first derived environmental factor and the second derived environmental factor in each re-cutting unit, reflecting the inherent environmental impact factors of the unit. For example, a re-cutting unit containing a concrete beam column (first derived environmental factor: poor thermal conductivity) and a cloth curtain (second derived environmental factor: high dust absorption rate) has a static environment set containing both factors.

[0087] The environmental monitoring importance is the importance of the re-cutting unit corresponding to the static environment set in environmental monitoring, which depends on the influence of the derived environmental factors on the environment. For example:

[0088] An evaluation index system is established to assign weights to the derived environmental factors (for example, the second derived environmental factor has a weight of 60%, because the decorative material is the main source of pollutants; the first derived environmental factor has a weight of 30%; and the regional function has a weight of 10%).

[0089] Scoring rules are set: a scoring standard (1-10 points) is set for each factor. For example, formaldehyde release rate > 0.03 mg / (m 2 ·h) is scored 8 points (high risk), and < 0.01 mg / (m 2 ·h) is scored 3 points (low risk).

[0090] In the regional function, the bedroom is scored 8 points and the storage room is scored 3 points.

[0091] The weighted sum (comprehensive score = ∑factor score x weight) is calculated according to the weights, and the importance level is divided according to the score:

[0092] High importance: score ≥ 7 points (e.g. children's room + plywood decoration);

[0093] Medium importance: 4-6 points (e.g. living room + ordinary latex paint);

[0094] Low importance: ≤ 3 points (e.g. storage room + no release material).

[0095] The rough installation height range is the installation height interval of the sensor in the vertical direction determined according to the environmental monitoring importance. For example, for a unit that mainly monitors the environment around the human body, the rough installation height range may be set to 1.5-2.0 meters.

[0096] The rough installation horizontal range is the installation area interval of the sensor in the horizontal direction determined according to the environmental monitoring importance. For example, for a unit close to a decorative piece that releases pollutants, the rough installation horizontal range may be set to a region 0.5-1.5 meters away from the decorative piece.

[0097] The factor distribution structure is the structure of the distribution form and density of each derived environmental factor in the cutting space. For example, in a cutting space, the distribution structure of the temperature and humidity derived environmental factor may exhibit the characteristics of dense in the window area and sparse in the corner area.

[0098] The per-layer factor dense trend is the direction and trend of the dense distribution of the derived environmental factor in the vertical direction (height layering). For example, the formaldehyde factor may exhibit a trend of dense in the 1.0-1.5 meter height layer and gradually sparse upward and downward in the vertical direction.

[0099] The per-column factor dense trend is the direction and trend of the dense distribution of the derived environmental factor in the horizontal direction (horizontal column segment). For example, the PM2.5 factor may exhibit a trend of dense near the door and window area and sparse away from the door and window area in the horizontal direction.

[0100] The to-be-installed height range is the installation height interval of the sensor in the vertical direction determined according to the per-layer factor dense trend in the factor distribution structure, which is more accurate than the rough installation height range. For example, based on the formaldehyde factor being dense in the 1.2-1.6 meter height layer, the to-be-installed height range is determined to be 1.2-1.6 meters.

[0101] The to-be-installed horizontal range is the installation area interval of the sensor in the horizontal direction determined according to the per-column factor dense trend in the factor distribution structure, which is more accurate than the rough installation horizontal range. For example, based on the PM2.5 factor being dense in the 0.3-1.0 meter area away from the door and window, the to-be-installed horizontal range is determined to be 0.3-1.0 meters.

[0102] The range negotiation processing is an interval comparison of the to-be-installed range and the rough installation range to analyze the overlap of the two. For example, the to-be-installed height range is 1.2-1.6 meters, and the rough installation height range is 1.0-1.8 meters. The range negotiation processing is to analyze the overlapping part (1.2-1.6 meters) of the two intervals.

[0103] The first range set is a set of height intervals that are most frequently negotiated with the to-be-installed height range in the rough installation height range. For example, the rough installation height range is 1.0-2.0 meters, and after multiple to-be-installed height ranges are negotiated with this range, the 1.4-1.6 meter interval overlaps the most frequently. This interval is the first range set.

[0104] The second range set is a set of horizontal intervals that are most frequently negotiated with the to-be-installed horizontal range in the rough installation horizontal range. For example, the rough installation horizontal range is 0-3 meters, and after multiple to-be-installed horizontal ranges are negotiated with this range, the 1.0-1.5 meter interval overlaps the most frequently. This interval is the second range set.

[0105] The intersection range segment is the interval segment of the overlap between the to-be-installed range and the rough installation range, for example, the intersection range segment of the to-be-installed horizontal range 0.5-1.2 meters and the rough installation horizontal range 0.3-1.5 meters is 0.5-1.2 meters.

[0106] The range overlap segment is the interval segment with the most overlaps between all to-be-installed ranges when the range set is empty, for example, the multiple to-be-installed height ranges are 1.1-1.3 meters, 1.2-1.4 meters, and 1.3-1.5 meters, and their range overlap segment is 1.3-1.4 meters (with the most overlaps).

[0107] The first final segment is the vertical height interval segment finally determined from the first range set for sensor installation, for example, the first range set is 1.3-1.6 meters, from which 1.3-1.5 meters is determined as the first final segment.

[0108] The vertical height is the total height of the cutting space in the vertical direction, and the first ratio is the ratio of the length of the first final segment to the vertical height.

[0109] The second final segment is the horizontal length interval segment finally determined from the second range set for sensor installation, for example, the second range set is 1.0-1.5 meters, from which 1.0-1.4 meters is determined as the second final segment.

[0110] The horizontal length is the total length of the cutting space in the horizontal direction, and the second ratio is the ratio of the length of the second final segment to the horizontal length.

[0111] The relative direction is the directional relationship (such as intersection, parallel, etc.) between the position of the first final segment in the vertical direction and the position of the second final segment in the horizontal direction, for example, the first final segment is located in the upper middle part of the vertical direction, and the second final segment is located in the middle part of the horizontal direction, and the two have an intersecting relative direction.

[0112] The key space is a space area determined based on the first final segment and the second final segment, which is crucial for environmental monitoring, and is an area where sensors need to be concentratedly deployed, for example, the first final segment is 1.3-1.5 meters (vertical direction), and the second final segment is 1.0-1.4 meters (horizontal direction), and the area formed by the intersection of the two is the key space.

[0113] The first concentrated distribution is a concentrated deployment mode of sensor groups in the key space to ensure accurate monitoring of the key area, for example, 2 sensor groups per square meter are deployed in the key space.

[0114] The second uniform distribution is an average deployment mode of sensor groups in the non-key space, for example, 1 sensor group per 5 square meters is deployed in the non-key space.

[0115] The beneficial effects of the above technical scheme are: through fine cutting of the building decoration environment, combining with the derived environmental factors of building and decoration to determine the sensor installation range, finally realizing the concentrated distribution of the sensor group in the key space and the uniform distribution in the non-key space, improving the effectiveness and pertinence of the sensor layer data collection, and laying a solid foundation for the accurate monitoring of the whole building decoration intelligent environment monitoring system.

[0116] The application provides a building decoration intelligent environment monitoring system based on an Internet of Things.

[0117] The coefficient determination unit is used for determining the cutting coefficient of the refined unit according to the building distribution and the decoration distribution of each cutting space.

[0118]

[0119] Wherein, U is the refined unit cutting coefficient; and is the building distribution influence weight coefficient of the corresponding cutting space. represents the building distribution complexity index, V is the volume of the i1th building component in the corresponding cutting space, V0 is the total volume of the corresponding cutting space, g is the environmental interference coefficient of the i1th building component in the corresponding cutting space, and n1 is the total number of building components in the corresponding cutting space; is the decoration distribution influence weight coefficient of the corresponding cutting space. is a decoration distribution sensitive index, S is the surface area of the i2th decoration component in the corresponding cutting space, S0 is the total surface area of the corresponding cutting space, f is the environmental release coefficient of the i2th decoration component in the corresponding cutting space; Fs is a space function correlation factor; Gk is a space communication interference coefficient; is the space topology weight coefficient of the corresponding cutting space; S1 is the effective ventilation area of the door and window of the corresponding cutting space; S2 is the total wall area of the corresponding cutting space; and fq is the air flow disturbance coefficient of the corresponding cutting space.

[0120] The re-cutting unit is used for rounding up the cutting coefficient to perform re-cutting on the corresponding cutting space.

[0121] In the embodiment, U is used for quantifying the fine degree of re-cutting of the cutting space, and the larger the value is, the more the number of refined units after cutting is, and the more detailed the division is. It is a core decision index of multiple factors such as building distribution, decoration distribution and space function.

[0122] Based on the building design drawings and past project experience, a mapping table of “building component type-influence weight” is established (such as a load-bearing wall∝=0.6, a common partition wall∝=0.3), and the value of is determined by manual or algorithm matching in combination with the actual building component composition of the cutting space.

[0123] Traverse all building components in the cutting space, get each component volume Vi1, total volume of cutting space V0 (measured by 3D modeling software), query the "building component-environment disturbance coefficient gi1" database (such as brick wall gi1 = 1.2, glass wall gi1 = 0.6), and substitute into the formula for cumulative calculation.

[0124] Referring to the environmental protection level of decorative materials and pollutant release characteristics, a "decoration type-influence weight" mapping table is established (such as plywood β = 0.6, latex paint β = 0.3), combined with the decoration configuration of the cutting space, automatically or manually matched to determine the β value.

[0125] Statistical surface area Si2 of all decorative pieces in the cutting space (extracted by 3D scanning or design drawings), total surface area S0 of the cutting space (calculated by 3D modeling), query "decoration material-environmental release coefficient fi2" database (such as solid wood fi2 = 1.5, cloth fi2 = 0.5), substitute into the formula for cumulative calculation.

[0126] According to the building space function classification standard (such as "Building Design Specification"), a "function type-correlation factor" table is established (such as bedroom Fs = 1.2, multi-purpose hall Fs = 1.8), and the Fs value is matched by identifying the functional attributes of the cutting space.

[0127] Based on the connectivity of the building space (closed, semi-open, open), "connectivity type-disturbance coefficient" rules are formulated (such as closed space Gk = 0, open type Gk = 0.8), combined with the actual connectivity state of the cutting space (analyzed by BIM model) to determine the Gk value.

[0128] Analyze the door and window layout and ventilation path of the cutting space, and establish a "topological complexity-weight coefficient" mapping (such as multiple doors and windows, complex air flow path γ = 0.3), combined with the actual topological structure to determine the γ value.

[0129] Simulate the air flow field of the cutting space by computational fluid dynamics (CFD), or establish a "air flow pattern-disturbance coefficient" table based on building ventilation design parameters (such as door and window orientation, spacing) (such as cross breeze fq = 1.3, stable air flow fq = 0.8), and match to determine the fq value.

[0130] Call the 3D space cutting algorithm (such as the space segmentation tool based on the BIM platform), input the number of refinement units corresponding to the cutting coefficient U, and the cutting boundary rules (derived from building distribution, decoration distribution, etc.), automatically generate the subspace model after the second cutting, and output the space range and attribute label of each refinement unit (such as sofa area-high decoration sensitivity).

[0131] The beneficial effects of the above technical scheme are: by constructing a cutting coefficient model of multi-dimensional parameter (building distribution, decoration distribution, space function, etc.) fusion, the refined requirements of cutting space are accurately quantified. The full-element correlation from building structure, decoration characteristics to space function is realized, the refined unit of re-cutting is not only suitable for environment monitoring requirements (such as fine cutting of high-pollution decoration area), but also adapts to space physical properties (such as reasonable segmentation of complex airflow area), which lays a solid foundation for subsequent accurate deployment of sensors and accurate monitoring of environmental quality, and significantly improves the spatial adaptability and intelligent level of the building decoration environment monitoring system.

[0132] The application provides a building decoration intelligent environment monitoring system based on an Internet of Things, and further comprises:

[0133] The starting module is configured to start the multi-modal acquisition of the building decoration area by the sensor layer and synchronously acquire the operation characteristic data of each sensor group.

[0134] The image analysis module is configured to acquire the initial thermal imaging image of each sensor group by the infrared sensor deployed in the building decoration area, and adjust the initial thermal imaging image to obtain a real-time thermal imaging image in combination with the temperature and humidity data.

[0135] The basic analysis module is configured to generate an analysis pair based on the operation characteristic data and the real-time thermal imaging image of each sensor group, periodically segment a series of analysis pairs collected in real time, and generate a basic feature set.

[0136] The difference analysis module is configured to predict a temperature change curve of the sensor group according to the operation characteristic data, and perform time alignment processing with the real-time thermal imaging image to obtain a temperature difference feature set.

[0137] The threshold setting module is configured to generate a transmission trigger threshold based on the basic feature set and the temperature difference feature set, divide all sensor groups into different grades according to the transmission trigger threshold, set the sensor groups after the division to be synchronously triggered according to the minimum trigger threshold under the corresponding grade, and transmit the collected data from the sensor layer to the data processing layer based on the Internet of Things communication layer.

[0138] In this embodiment, multi-modal acquisition is the synchronous acquisition of multiple environmental parameters (temperature and humidity, formaldehyde concentration, and different physical / chemical modalities), which comprehensively reflects the state of the building decoration environment. For example, in the office area, the sensor layer simultaneously acquires temperature (physical modality), carbon dioxide concentration (chemical modality), and illumination (optical modality), covering multi-dimensional information such as thermal environment, air quality, and light environment. Specifically, multiple types of sensors (temperature and humidity sensor selects SHT31, formaldehyde sensor selects ZE08-CH2O) are configured in the sensor layer, a collection control program is written, and the collection actions of different sensors are triggered synchronously at fixed time intervals (such as every 1 minute). The multi-modal data is cached to the local storage of the sensor.

[0139] Sensor group operating characteristic data is a parameter reflecting the working state of the sensor group itself, including voltage, current, signal frequency, energy consumption, fault code, etc., which is used to monitor whether the sensor is running normally. For example, the operating characteristic data of a certain temperature and humidity sensor group is: working voltage 3.3V, signal transmission frequency 9600Hz, and no fault code for 3 hours of continuous operation. If the voltage drops to 3.0V, it may indicate that the sensor power supply is abnormal. Specifically, the operating parameters are read from the control module (such as MCU chip) of the sensor, and the data is obtained in real time through I 2 C, SPI, etc. or embedded in the sensor firmware to report data to the start module at regular intervals.

[0140] The infrared sensor is a sensor that captures the temperature distribution of the surface of an object using infrared thermal radiation principles, and outputs a thermal imaging image, which is used to monitor the thermal state of the sensor group. For example, in a smart home, an infrared sensor deployed on the ceiling can take a thermal imaging image of the sensor group in the living room area, showing the temperature change of the temperature and humidity sensor due to long-term operation. Specifically, the resolution of the infrared sensor (such as 160x120 pixels) is selected according to the monitoring range (such as a 10m x 10m area), and the sensor is installed on the top of the building or embedded in the wall through a bracket or embedded in the wall. The communication protocol (such as WiFi, USB) of the sensor is configured to access the system.

[0141] In this embodiment, the initial thermal imaging image is an uncorrected image directly collected by the infrared sensor, which contains interference information such as environmental temperature and humidity, sensor error, etc. For example, in a low-temperature environment in winter, the initial thermal imaging image taken by the infrared sensor may show a temperature of the sensor group that is 2°C lower than the actual value due to the low environmental temperature. Specifically, the infrared sensor collects images at a fixed frame rate (such as 5 frames / second) and temporarily stores them in raw data format (such as 16-bit grayscale images). The images are transmitted to the image analysis module through a network or data bus and are waiting for subsequent correction.

[0142] In this embodiment, the temperature and humidity data are air temperature and relative humidity data collected by the sensor layer, which are used to correct the environmental interference of the infrared thermal imaging image. For example, if the temperature and humidity data of a certain area are 25°C and 50%, the image analysis module uses the data to correct the problem of infrared image thermal radiation attenuation caused by high humidity. Specifically, the data are obtained in real time from the temperature and humidity sensor (such as DHT11) of the sensor layer and transmitted to the image analysis module through MQTT, HTTP, or other protocols as input parameters for image correction.

[0143] In this embodiment, the real-time thermal imaging image is an image that can accurately reflect the actual temperature distribution of the sensor group after the initial thermal imaging image is corrected by combining the temperature and humidity data, which is used for subsequent analysis. For example, the corrected real-time thermal imaging image clearly shows that the temperature of a certain sensor group is 10°C higher than normal due to a circuit short circuit, which helps to quickly locate the fault. Specifically, a “temperature and humidity-infrared correction model” (such as a temperature compensation algorithm based on Planck's law) is established, the initial thermal imaging image and the temperature and humidity data are input, and the real-time thermal imaging image is generated through matrix operation, gray correction, and other algorithms. The image is stored in a standard image format (such as PNG) for subsequent module calls.

[0144] In this embodiment, the analysis pair is a pair of associated data composed of the operating characteristic data of a single sensor group and the corresponding real-time thermal imaging image, which is used to analyze the “operating state-thermal performance” of the sensor. For example, the analysis pair of a certain formaldehyde sensor group includes its working current (operating characteristic data) and the temperature distribution area of the sensor group in the thermal imaging image (real-time thermal imaging image). Specifically, a unique identifier is generated for each sensor group by matching the timestamp and spatial coordinates, the operating characteristic data and the real-time thermal imaging image cutting area under the same identifier are associated to form an analysis pair, and the analysis pair is stored in the association table of the database.

[0145] In this embodiment, the period segmentation is the division of the continuous analysis pair sequence according to a fixed time period (such as 1 hour, 1 day), and the characteristic rules in different periods are extracted. For example, the analysis pair of the office sensor group is divided into “working day 9:00-18:00” and “non-working day” periods, and the characteristics of the high-load operation of the sensor during the working period are extracted. Specifically, a period configuration file (such as JSON format, defining the period length and start time) is set in the basic analysis module, the analysis pair sequence is filtered by timestamp, and the continuous data is divided into multiple period subsequences, each of which corresponds to an analysis period.

[0146] Temperature change curve prediction is to predict the temperature change trend of the sensor group over time by machine learning or physical model using sensor group operation characteristic data (such as voltage, current), and output a curve, for example, based on the historical working current data of a certain temperature and humidity sensor group, an LSTM model is trained to predict the temperature change curve of the sensor group from 30°C to 35°C linearly in the next 1 hour. Specifically, the historical operation characteristic data of the sensor group and the corresponding temperature (extracted from the thermal imaging image) are collected to build a training data set, a prediction model (such as linear regression, LSTM) is selected for training, and the model is deployed in the difference analysis module to input real-time operation characteristic data and output temperature change curve.

[0147] In this embodiment, the time alignment processing is to match the predicted temperature change curve with the real-time thermal imaging image in the time dimension, to ensure the comparison of data at the same time point and eliminate the difference error caused by time offset, for example, the timestamp of the predicted temperature curve is "10:00-11:00", and the shooting time of the real-time thermal imaging image is 10:30. Through time interpolation or truncation, the two are aligned to the 10:30 time point for comparison. Specifically, the timestamps of the predicted curve and the thermal imaging image are recorded, a linear interpolation algorithm (such as Python's scipy.interpolate) is used to correct the time of the predicted curve, or the thermal imaging image sequence is cropped to ensure that the time dimensions of the two are consistent, and then the difference is calculated.

[0148] In this embodiment, the transmission trigger threshold is a critical value for determining whether the sensor group data needs to be transmitted immediately to the data processing layer. The lower the threshold, the more priority the data has for transmission. For example, the transmission trigger threshold of a certain sensor group is 0.8 (after normalization), and when the real-time characteristic calculation value of the sensor group exceeds 0.8, the data is triggered for emergency transmission.

[0149] Sensor group level division is to divide the sensor group into different priority levels (such as emergency, ordinary, and low priority) according to the transmission trigger threshold, to determine the data transmission strategy. If the threshold <0.5 is set as the emergency level (immediate transmission), 0.5-0.8 is set as the ordinary level (timed transmission), and >0.8 is set as the low priority level (batch transmission), a sensor group with a threshold of 0.6 is classified as an ordinary level.

[0150] Synchronous trigger setting is to configure the minimum trigger threshold of the corresponding level according to the level of the sensor group, to ensure that the sensor groups of the same level trigger data transmission according to a unified strategy. Specifically, the minimum trigger threshold of each level is preset (such as supplementing the threshold field in the level division rule table), and the threshold setting module automatically loads the minimum trigger threshold of the corresponding level after determining the level of the sensor group. The trigger configuration instruction is sent to the sensor group through the Internet of Things communication layer to set the trigger condition.

[0151] In the Internet of Things communication layer configuration data transmission protocol (such as MQTT, CoAP), set up different levels of transmission queue (emergency queue, ordinary queue), threshold setting module triggers data transmission, sensor layer will cache data according to protocol encapsulation, send to the receiving server of data processing layer, data processing layer analysis and storage.

[0152] The beneficial effects of the above technical solutions are: through the starting module synchronous triggering multi-modal environment acquisition and sensor state monitoring, the image analysis module combines with temperature and humidity to correct the thermal imaging image, the basic analysis and difference analysis module extracts features in layers, and the threshold setting module dynamically controls data transmission, which constructs an intelligent collaborative mechanism of the whole process of acquisition-processing-transmission. It not only guarantees the comprehensiveness of environmental parameter acquisition, but also improves data quality through sensor running state monitoring and thermal imaging correction, realizes differentiated and accurate triggering of data transmission, provides an efficient and reliable front-end data processing and uploading scheme for the building decoration intelligent environment monitoring system, and effectively supports subsequent environmental quality modeling and early warning decision-making.

[0153] The application provides a building decoration intelligent environment monitoring system based on Internet of Things.

[0154] The pulse acquisition unit is used for acquiring historical interruption time length of each sensor group in a working process and each two continuous measurement pulses before and after the historical interruption time length, and comparing time widths of the first two continuous measurement pulses and the last two continuous measurement pulses to obtain a basic error correction width.

[0155] The clock deviation determination unit is used for synchronously acquiring clock deviation data of each sensor device of each sensor group, calculating frequency compensation of a device clock of each sensor relative to a reference clock of the sensor group, and constructing a clock deviation sequence of the sensor group.

[0156] The symbol determination unit is used for associating and comparing the basic feature set and the temperature difference feature set to obtain an initial triggering threshold, and according to a direction pair of the basic feature set and the temperature difference feature set, and in combination with the clock deviation sequence, a maximum deviation and a minimum deviation in the clock deviation sequence, the judgment symbol is given to the basic error correction width.

[0157] The threshold acquisition unit is used for obtaining a transmission triggering threshold according to the initial triggering threshold, the judgment symbol and the error correction width.

[0158] In this embodiment, the historical interruption duration is the length of time that the sensor group stops data collection or transmission due to various reasons (such as communication failure, temporary hardware failure, etc.) during operation. Specifically, in the control program of the sensor group, an interruption monitoring module is set up. When an abnormal stop of data transmission or collection is detected, a timer is started. When normality is restored, the timer is stopped. The recorded time difference is the historical interruption duration.

[0159] Continuous measurement pulses are electrical pulse signals for transmitting data or indicating the working state of the sensor group, which are continuously generated according to a certain rule during normal operation. Taking a temperature and humidity sensor group as an example, data is collected every 1 minute and transmitted in the form of pulses. The pulses of adjacent two data transmissions are continuous measurement pulses.

[0160] Time width comparison is to compare the time width of the first two continuous measurement pulses with the time width of the last two continuous measurement pulses. The basic error correction width is the difference obtained by comparing the time width of the first and last continuous measurement pulses, which is used for subsequent error correction and threshold calculation. If the average width of the first two pulses is 10 ms and the average width of the last two pulses is 8 ms, the basic error correction width is 2 ms.

[0161] Sensor device is a single sensor individual that constitutes the sensor group, such as a single temperature and humidity sensor, a single formaldehyde sensor, etc.

[0162] Clock deviation data is the time difference between the clock of a single sensor device and the reference clock of the sensor group. For example, if the reference clock of the sensor group shows 10:00:00 and the clock of a certain sensor device shows 10:00:02, the clock deviation data is +2 seconds. Specifically, in the sensor device and the reference clock module, high-precision clock circuits are configured. Through regular time synchronization communication (such as a simplified version of NTP protocol), the time difference between the two is calculated and recorded as clock deviation data.

[0163] Device clock is a clock system used by a sensor device for timing, which determines the time reference for internal data collection, processing and transmission of the sensor. In the hardware design of the sensor device, a clock chip or circuit (such as a crystal oscillator combined with a clock circuit) is integrated. The initial value, frequency and other parameters of the clock are configured through software.

[0164] Reference clock is a high-precision clock that serves as a time reference in the sensor group, which is used to calibrate the clocks of other sensor devices in the group.

[0165] Frequency compensation is to keep the device clock synchronized with the reference clock, the value of the frequency adjustment of the device clock, if the device clock is slow, increase the frequency compensation to make it faster, otherwise, reduce the frequency compensation. Specifically, according to the clock deviation data, the frequency value to be adjusted is calculated, the frequency of the device clock is modified through the clock control interface of the sensor device (such as the frequency adjustment pin of some clock chips) or the software configuration register, and the frequency compensation is realized.

[0166] The clock deviation sequence is a series of clock deviation data arranged in chronological order, which can reflect the change of clock deviation with time. For example, record the clock deviation data every 1 hour, and record continuously for one day. The sequence composed of these data in chronological order is the clock deviation sequence.

[0167] The basic feature set is a feature set reflecting the basic working state of the sensor group after the original data collected by the sensor group is preliminarily processed (such as filtering, normalization, etc.). For example, the average temperature, humidity fluctuation amplitude and other features obtained by processing the temperature and humidity data collected by the temperature and humidity sensor group form a set.

[0168] The temperature difference feature set is a feature set reflecting the temperature difference between the predicted temperature change curve of the sensor group and the actual thermal imaging image temperature. For example, by comparing the temperature change curve obtained by the prediction model with the actual thermal imaging measured temperature, the temperature deviation mean, maximum deviation and other features form a set.

[0169] The initial trigger threshold is a threshold obtained by correlating and comparing the basic feature set and the temperature difference feature set (such as using machine learning algorithm, statistical analysis method), which is used to judge whether to trigger data transmission and other operations. For example, a linear regression model is used to analyze the relationship between the basic feature set and the temperature difference feature set, and the initial trigger threshold is calculated. Specifically, select a suitable data analysis algorithm (such as support vector machine, linear regression), write the algorithm program in the software, input the basic feature set and the temperature difference feature set data, train the model and calculate the initial trigger threshold.

[0170] The direction pair is the relationship between the change trend of the basic feature set and the temperature difference feature set, including the same direction (both rising or both falling) and the opposite direction (one rising and one falling). If the temperature mean in the basic feature set rises, and the temperature deviation in the temperature difference feature set also rises, it is the same direction of the direction pair. Specifically, write a trend analysis algorithm in the software, calculate the change rate (such as the difference between the current value and the historical value divided by the time) of the basic feature set and the temperature difference feature set data, and judge the direction pair according to the sign of the change rate.

[0171] The judgment sign is the positive and negative sign of the basic correction width, which is used to represent the directional property of the deviation corresponding to the basic correction width in the overall data relationship, such as Table 1.

[0172] Table 1 determination rule of symbol

[0173]

[0174]

[0175] For example, the basic feature set (voltage fluctuation amplitude) rises (change rate + 0.2), the temperature difference feature set (temperature deviation mean) rises (change rate + 0.3), the direction pair is "same direction"; the clock deviation sequence trend is "deviation absolute value increases" (from -3s→-5s), and the current deviation -5s is close to the maximum deviation -6s (difference <20%). In line with "same direction + deviation increase + close to maximum deviation", the symbol is "+", indicating that the deviation corresponding to the basic error correction width is expanding in the adverse direction.

[0176] The transmission trigger threshold is the critical value finally determined for judging whether to trigger the sensor group data transmission to the data processing layer. When the feature value corresponding to the sensor group monitoring data exceeds the threshold, the data transmission is triggered, and the transmission trigger threshold = initial trigger threshold + symbol × basic error correction width × adjustment coefficient, and the value of the adjustment coefficient is generally 0.1.

[0177] The beneficial effects of the above technical scheme are: the pulse acquisition unit accurately captures the sensor group pulse signal deviation, the clock deviation determination unit guarantees time synchronization, the symbol determination unit determines the deviation direction attribute, the threshold acquisition unit fuses multi-dimensional data to calculate the transmission trigger threshold, which can effectively improve the accuracy and timeliness of the data transmission of the building decoration intelligent environment monitoring system, and provide reliable data transmission guarantee for subsequent environmental quality analysis and improvement decision.

[0178] The application provides a building decoration intelligent environment monitoring system based on Internet of Things.

[0179] The initial set determination module is used for single-parameter curve feature extraction on the preprocessed environmental parameter data, captures the slope mutation point, continuous abnormal interval and periodic fluctuation amplitude of each curve through an adaptive sliding window algorithm, and generates a single-parameter initial change set.

[0180] The mutation set determination module is used for calculating the dynamic cross-correlation coefficient of any two environmental parameter curves, screening the parameter pairs with correlation coefficients exceeding the threshold, and integrating the cooperative change features of the corresponding curves into the initial change set to form a prominent state change set fusing correlation information.

[0181] Preferably, the data processing layer further comprises:

[0182] The segment analysis module is configured to perform curve segmentation on the preprocessed data, divide analysis subspaces according to the functional characteristics and material characteristics of the architectural decoration region, and extract significant change nodes of each environmental parameter curve in each subspace through an extreme point clustering algorithm.

[0183] The subspace set determination module is configured to, for each subspace, establish a mapping relationship among material, structure and environmental parameters in combination with physical attribute parameters of the decoration material and spatial structure parameters, filter interference changes in the subspace through the mapping relationship, retain prominent state changes that are strongly related to decoration characteristics, and form a subspace-specific change set.

[0184] The model determination module is configured to construct a hierarchical structure of a main model and sub-models based on the subspace-specific change set and a required change set extracted from the prominent state change set, and form a dynamically adaptive architectural decoration environmental quality model.

[0185] Preferably, the main model learns common environmental quality rules across spaces, and the sub-models perform parameter fine-tuning for decoration scenarios.

[0186] In this embodiment, single-parameter curve feature extraction is performed on a time series curve of a single environmental parameter (such as a formaldehyde concentration curve over time), and features (such as mutation points and abnormal intervals) reflecting the change rule are extracted. For example, for a temperature curve, a slope mutation point of “15:00 temperature suddenly rising from 25°C to 30°C” and a continuous abnormal interval of “humidity > 70% for 3 hours” are extracted.

[0187] The adaptive sliding window algorithm is an analysis algorithm in which the window size can be automatically adjusted according to data changes. When data fluctuates greatly, the window size is reduced (to increase sensitivity), and when data fluctuates slightly, the window size is expanded (to reduce false positives).

[0188] A slope mutation point is a point at which the slope of an environmental parameter curve suddenly increases or decreases, reflecting a dramatic change in the parameter in a short period of time. For example, after the ventilation system is started, the CO2 concentration curve decreases from 1200 ppm to 800 ppm in 1 minute, and the corresponding point is a slope mutation point. Specifically, the slope of adjacent points of a curve (for example, the concentration change amount of a formaldehyde concentration curve divided by the time difference between t1 and t2) is calculated, and when the absolute value of the slope exceeds a threshold value (for example, 0.02 mg / m 3 / min), the point is marked as a mutation point.

[0189] A continuous abnormal interval is an interval in which an environmental parameter exceeds the normal range and lasts for a certain period of time, which is different from transient fluctuations.

[0190] Periodic fluctuation amplitude is the amplitude of periodic variation in the environmental parameter curve (such as the fluctuation range of daily morning and evening temperature), reflecting the strength of regular variation of the parameter, for example, the bedroom temperature in summer presents a cycle of “24℃ (early morning) → 30℃ (afternoon) → 25℃ (night)”, and the fluctuation amplitude is 6℃.

[0191] Single-parameter initial variation set is a set composed of slope mutation points, continuous abnormal intervals, and periodic fluctuation amplitudes of a single environmental parameter, reflecting the independent variation characteristics of the parameter, for example, the initial variation set of temperature and humidity parameters contains “3 slope mutation points, 1 continuous high-temperature interval, and a daily fluctuation amplitude of 5℃”.

[0192] Dynamic cross-correlation coefficient is a correlation index (range -1 to 1) that measures the correlation between two environmental parameter curves at different time lags, and the closer the value is to 1, the stronger the correlation, for example, when the temperature curve lags behind the formaldehyde concentration curve by 30 minutes, the cross-correlation coefficient is 0.7, indicating that the formaldehyde concentration rises significantly after the temperature rises by 30 minutes, and the two are dynamically correlated, specifically: for two parameter curves (such as temperature T and formaldehyde F), calculate the correlation coefficient (corr(T(t), F(t+τ))) at different lag times τ, and take the maximum value as the dynamic cross-correlation coefficient.

[0193] Parameter pair is a combination of any two environmental parameters (such as “temperature-formaldehyde” and “humidity-PM2.5”), used to analyze the correlation between parameters, for example, the dynamic cross-correlation coefficient of the “temperature-formaldehyde” parameter pair is 0.6 (above the threshold value of 0.5), which is selected as a strongly correlated parameter pair.

[0194] Synergistic variation feature is the synchronous or lagging variation feature (such as co-directional rise or reverse fluctuation) exhibited by the curves of a strongly correlated parameter pair, for example, the synergistic variation feature of the “temperature-formaldehyde” parameter pair is “for every 1℃ rise in temperature, the formaldehyde concentration rises by 0.02 mg / m 3 ” after a 10-minute lag”.

[0195] Prominent state variation set is a set formed by integrating the single-parameter initial variation set and the synergistic variation feature of the strongly correlated parameter pair, reflecting the environmental prominent variation under multi-parameter correlation, for example, the prominent state variation set contains “formaldehyde single-parameter mutation point, temperature single-parameter abnormal interval, and temperature-formaldehyde synergistic rise feature”.

[0196] Curve segmentation is the splitting of the overall curve of an environmental parameter into local curves in each subspace, facilitating targeted analysis, for example, the overall curve of the living room PM2.5 concentration is split into “sofa area local curve” and “balcony area local curve”, reflecting the pollution conditions in different subspaces.

[0197] Function property is the usage function of the decoration area (e.g. “rest” for bedroom, “cooking” for kitchen), which determines the monitoring focus of the environmental parameters.

[0198] Material property is the physical and chemical attributes of the decoration materials (e.g. “low formaldehyde release” for solid wood, “high formaldehyde release” for plywood), which affects the change pattern of the environmental parameters.

[0199] Analysis sub-space is the local space unit divided by function property and material property, which is the minimum space unit for environmental analysis. For example, the master bedroom is divided into “bed area (solid wood floor + fabric bed)” and “closet area (plywood cabinet) ”.

[0200] Extreme point clustering algorithm is to cluster the local extreme points (peak, valley) in the environmental parameter curve according to the similarity, and extract the representative significant change nodes. K-means clustering algorithm (K value is set to 3-5) is used to cluster the extreme points, and the center value of each class is taken as the significant change node.

[0201] Significant change node is the representative extreme point or trend turning point in the environmental parameter curve in the sub-space, which reflects the key environmental change of the sub-space. For example, the formaldehyde concentration curve in the closet area reaches a peak of 0.18 mg / m 3 3 days after the new closet is installed, which is the significant change node.

[0202] Physical property parameter of decoration material is the inherent attribute of decoration material (e.g. air permeability, pollutant release rate, adsorption rate), which affects the change of environmental parameters. For example, the physical property parameters of solid wood floor are “formaldehyde release rate 0.01 mg / (m 2 ·h), air permeability 0.8 (high)”; and the physical property parameters of plywood are “release rate 0.05 mg / (m 2 ·h), air permeability 0.3 (low)”.

[0203] Space structure parameter is the structural characteristics of the sub-space (e.g. ventilation coefficient, volume, number of doors and windows), which affects the diffusion and accumulation of environmental parameters. For example, the space structure parameters of the closet area are “ventilation coefficient 0.2 (low, due to closure), volume 2 m 3 ”; and the space structure parameters of the central area of the living room are “ventilation coefficient 0.8 (high, due to proximity to windows), volume 30 m 3 ”.

[0204] Material-structure-environment parameter mapping relationship is to describe the correlation between the physical properties of decoration materials, space structure parameters and the change of environmental parameters (e.g. “high release material + low ventilation → formaldehyde easy to accumulate”), such as “plywood (release rate 0.05) + ventilation coefficient 0.2 → formaldehyde concentration increases by 0.03 mg / m 3”, specifically: training a machine learning model (such as random forest) with experimental data, inputting material properties and structural parameters, and outputting environmental parameter change rules to form a mapping relationship.

[0205] Interference changes are environmental parameter changes within the subspace that are unrelated to decoration characteristics (such as fluctuations caused by short-term ventilation, personnel activities), which need to be filtered to avoid interference analysis, for example, the wardrobe area has a 0.02 mg / m 3 drop in formaldehyde concentration due to temporary door ventilation (not caused by material release), which belongs to interference changes.

[0206] The subspace-specific change set is the set of prominent state changes that are strongly related to decoration material characteristics and space structure after filtering out interference changes within the subspace, for example, the wardrobe area-specific change set includes "persistent formaldehyde increase nodes caused by plywood release, concentration accumulation characteristics under closed structure".

[0207] The required change set is the change feature related to a specific subspace extracted from the prominent state change set (such as extracting changes only belonging to the wardrobe area from the overall prominent change set), for example, from the overall prominent state change set in the living room, extract "formaldehyde peak change only in the wardrobe area" as the required change set for this subspace.

[0208] The main model is a basic model that learns the common environmental quality rules across subspaces, suitable for general analysis of all subspaces, for example, the main model learns the common rule that "in all spaces, a 10℃ temperature increase will increase the release rate of formaldehyde from decorative materials by 20%".

[0209] The sub-model is a model that fine-tunes the main model parameters for a specific decoration scenario (such as a child's room, kitchen), adapting to the scene's special nature, for example, the child's room sub-model lowers the formaldehyde concentration warning threshold from 0.1 mg / m 3 to 0.08 mg / m 3 , adapting to the high sensitivity of children to formaldehyde.

[0210] The hierarchical structure is the nested structure of the main model and the sub-model, with the main model providing the basic framework and the sub-model optimizing for the scene, achieving a collaborative analysis of commonality and individuality.

[0211] The dynamically adapted building decoration environmental quality model is a model that can automatically adjust the analysis logic according to the subspace decoration characteristics (materials, functions), allowing precise evaluation of environmental quality in different scenarios, for example, the model focuses on PM2.5 and CO2 when analyzing the living room, and focuses on formaldehyde and temperature and humidity when analyzing the bedroom, with parameter thresholds dynamically adjusted according to material changes.

[0212] The beneficial effects of the above technical scheme are: the initial set determination module captures single parameter change, the mutation set determination module fuses multi-parameter correlation characteristics, the segmented analysis module realizes spatial fine analysis, the subspace set determination module filters interference and focuses on the change of decoration characteristics, and the model determination module constructs a hierarchical model, forming a complete link from data extraction to model construction. The problem of insufficient adaptation of traditional models to spatial heterogeneity and decoration characteristics is effectively solved.

[0213] Obviously, those skilled in the art can make various modifications and variations to the present application without departing from the spirit and scope of the present application. Thus, if these modifications and variations of the present application fall within the scope of the claims of the present application and their equivalent technologies, the present application also intends to include these modifications and variations.

Claims

1. An Internet of Things based intelligent environment monitoring system for interior decoration, characterized in that, The application relates to a sensor layer deployed in a building decoration area, which is used for collecting environmental parameters in a building decoration environment in real time, including temperature and humidity, formaldehyde concentration, PM2.5 concentration, illumination and carbon dioxide concentration. An Internet of Things communication layer is used for transmitting data collected by the sensor layer to a data processing layer according to a transmission standard of the sensor layer. The data processing layer is used for pre-processing real-time data collected by the sensor, including data filtering, abnormal value detection and data normalization processing, analyzing and determining a prominent state change set of each environmental parameter based on a data curve based on the pre-processed data, and modeling to establish a building decoration environment quality model. The data processing layer is also used for analyzing the building decoration environment quality model according to a preset environmental quality standard to generate an environmental monitoring report and early warning information, and delivering the report and the information to an application layer control execution device for operation to improve environmental quality. The application further relates to a cutting module used for cutting a three-dimensional area structure of the building decoration environment according to area functions to obtain a plurality of cutting spaces, and determining a refined unit for secondary cutting according to building distribution and decoration distribution of each cutting space.

2. The IoT based interior decoration smart environment monitoring system as claimed in claim 1, wherein, A set construction module is used for generating a static environment set of a secondary cutting unit according to a first derived environmental factor of a building type and a second derived environmental factor of a decoration type involved in each secondary cutting unit in a same cutting space. A rough determination module is used for analyzing environmental monitoring importance of each static environment set, and locking a rough installation height range and a rough installation horizontal range according to the environmental monitoring importance. A single installation range determination module is used for constructing a factor distribution structure of each derived environmental factor in a corresponding cutting space based on all static environment sets in the same cutting space, analyzing each layer factor dense trend and each column factor dense trend in the factor distribution structure, and locking a to-be-installed height range and a to-be-installed horizontal range corresponding to the corresponding derived environmental factor. A negotiation processing module is used for range negotiation processing of all to-be-installed height ranges and to-be-installed horizontal ranges with the rough installation height range and the rough installation horizontal range respectively, determining a first range set with the most negotiation times in the rough installation height range and a second range set with the most negotiation times in the rough installation horizontal range, wherein the range set is empty or contains at least one negotiation range segment, and when the range set is empty, a range overlap segment with the highest number of corresponding all to-be-installed ranges is determined. A distribution module is used for determining a first concentrated distribution of a sensor group for a key space according to a first final segment in a vertical direction and a first ratio of a vertical height, a second final segment in a horizontal direction and a second ratio of a horizontal length, and combining the relative direction of the first final segment and the second final segment, and simultaneously performing a second uniform distribution of the sensor group in a non-key space, wherein the key space is determined based on the first final segment and the second final segment, and the non-key space is a remaining space in the corresponding cutting space except the key space. The distribution result of the sensor group is used as the sensor layer deployed in the building decoration area. The cutting module comprises: ​ 3. The IoT based interior decoration smart environment monitoring system as claimed in claim 2, wherein, ​ The coefficient determination unit is configured to determine the cutting coefficient of the refined unit according to the building distribution and the decoration distribution of each cutting space; wherein, U is the refined unit cutting coefficient; and a is the building distribution influence weight coefficient corresponding to the cutting space; represents the building distribution complexity index, Vii is the volume of the ith building component in the corresponding cutting space, Vo is the total volume of the corresponding cutting space, gii is the environmental interference coefficient of the ith building component in the corresponding cutting space, and n1 is the total number of building components in the corresponding cutting space; and β is the decoration distribution influence weight coefficient corresponding to the cutting space; is the decoration distribution sensitive index, Sii is the surface area of the ith decoration component in the corresponding cutting space, So is the total surface area of the corresponding cutting space, fi2 is the environmental release coefficient of the ith decoration component in the corresponding cutting space; Fs is the space function correlation factor; Gk is the space connection interference coefficient; γ is the space topology weight coefficient corresponding to the cutting space; S1 is the effective ventilation area of the door and window of the corresponding cutting space; S2 is the total wall area of the corresponding cutting space; and fq is the air flow disturbance coefficient of the corresponding cutting space; The re-cutting unit is configured to round up the cutting coefficient to re-cut the corresponding cutting space.

4. The IoT based interior decoration smart environment monitoring system as claimed in claim 1, wherein, Further comprising: The starting module is configured to start the sensor layer to collect multi-modal collection of the building decoration area, and synchronously collect operation characteristic data of each sensor group; The image analysis module is configured to collect initial thermal imaging images of each sensor group by the infrared sensor arranged in the building decoration area, and adjust the initial thermal imaging images to obtain real-time thermal imaging images in combination with temperature and humidity data; The basic analysis module is configured to generate analysis pairs based on the operation characteristic data of each sensor group and the real-time thermal imaging images, periodically divide a series of analysis pairs collected in real time, and generate a basic feature set; The difference analysis module is configured to predict a temperature change curve of the sensor group according to the operation characteristic data of the sensor group, and perform time alignment processing with the real-time thermal imaging images to obtain a temperature difference feature set; The threshold setting module is configured to generate a transmission trigger threshold based on the basic feature set and the temperature difference feature set, divide all sensor groups according to the transmission trigger threshold, set the divided sensor groups to be synchronously triggered according to the minimum trigger threshold under the corresponding level, and transmit the collected data from the sensor layer to the data processing layer based on the Internet of Things communication layer.

5. The IoT based interior decoration smart environment monitoring system as claimed in claim 4, wherein, The threshold setting module comprises: The pulse acquisition unit is configured to acquire historical interruption time lengths of each sensor group in the working process, and acquire each two continuous measurement pulses before and after the historical interruption time lengths, compare time widths of the first two continuous measurement pulses and the last two continuous measurement pulses, and obtain a basic error correction width; The clock deviation determination unit is configured to synchronously collect clock deviation data of each sensor device of each sensor group, calculate frequency compensation of each sensor device clock relative to a reference clock of the sensor group, and construct a clock deviation sequence of the sensor group; The symbol determination unit is configured to compare and analyze the basic feature set and the temperature difference feature set to obtain an initial trigger threshold, assign a judgment symbol to the basic error correction width according to a direction pair of the basic feature set and the temperature difference feature set, in combination with the clock deviation sequence, a maximum deviation in the clock deviation sequence, and a minimum deviation; The threshold acquisition unit is configured to obtain a transmission trigger threshold according to the initial trigger threshold, the judgment symbol, and the error correction width.

6. The IoT based interior decoration smart environment monitoring system as claimed in claim 1 wherein, The data processing layer comprises: The initial set determination module is configured to perform single-parameter curve feature extraction on the preprocessed environmental parameter data, capture a slope mutation point, a continuous abnormal interval, and a periodic fluctuation amplitude of each curve by using a self-adaptive sliding window algorithm, and generate a single-parameter initial change set; The mutation set determination module is configured to calculate a dynamic cross-correlation coefficient of any two environmental parameter curves, filter parameter pairs with a correlation coefficient exceeding a threshold, include cooperative change features of corresponding curves into the initial change set, and form a prominent state change set with fusion correlation information.

7. The IoT based interior decoration smart environment monitoring system as claimed in claim 6, wherein, The data processing layer further comprises: The segmented analysis module is configured to perform curve segmentation on the preprocessed data, divide analysis subspaces according to functional characteristics and material characteristics of the architectural decoration region, and extract significant change nodes of each environmental parameter curve in each subspace through an extreme point clustering algorithm. The subspace set determination module is configured to, for each subspace, establish a mapping relationship among material, structure and environmental parameters in combination with physical attribute parameters of the decoration material and spatial structure parameters, filter interference changes in the subspace through the mapping relationship, retain prominent state changes that are strongly related to decoration characteristics, and form a subspace-specific change set. The model determination module is configured to construct a hierarchical structure of a main model and a submodel based on the subspace-specific change set and a required change set of the subspace extracted from the prominent state change set, and form a dynamically adaptive architectural decoration environmental quality model.

8. The IoT based interior decoration smart environment monitoring system as claimed in claim 7, wherein, The main model learns common environmental quality rules across spaces, and the submodel performs parameter fine-tuning for decoration scenes.

Images