A snap-on rapid deployment device for green roofs and a self-calibration method

Abstract

The application discloses a buckle type rapid laying equipment and a self-calibration method of a green roof, and relates to the technical field of green buildings.The application can be rapidly deployed and load reduction, modular design and buckle type interlocking structure make the installation process as simple as assembling building blocks, greatly shorten the construction period, and reduce the dependence on professional skills.The load burden on the roof structure is effectively reduced by the light tray unit and the cultivation substrate.The application can intelligently and accurately manage water, through an integrated sensor network and a self-calibration algorithm, the system can realize real-time sensing of the real moisture condition of the substrate, and drive the irrigation system to accurately supplement water, avoid water resource waste or insufficient water supply caused by traditional fixed time sequence irrigation, and improve water saving benefit and plant survival rate.

Description

Technical Field

[0001] This invention relates to the field of green building and rooftop greening technology, specifically to a snap-on rapid deployment device and self-calibration method for green roofs. Background Technology

[0002] Green roof systems, as an important technology in green building and sponge city construction, can effectively reduce building energy consumption, mitigate the urban heat island effect, promote rainwater management, and enhance biodiversity. However, traditional green roof systems still face many challenges in practical application:

[0003] Complex construction and high load requirements: Traditional green roofs usually require the on-site laying of waterproofing, drainage, filtration, and planting soil layers, etc., resulting in multiple structural layers, long construction period, and high requirements for the load-bearing capacity of the roof.

[0004] Low modularity and difficult maintenance and replacement: The non-modular design makes it inconvenient to repair when there is local damage, and may require large-scale overhaul, which increases maintenance costs and difficulty.

[0005] Insufficient level of intelligent irrigation management: Many systems lack effective monitoring methods, and irrigation decisions rely heavily on fixed time sequences or manual judgment, making it difficult to make precise adjustments based on the actual water needs of plants and environmental changes, which may lead to water waste or poor plant growth.

[0006] Lack of system status monitoring and calibration: Existing systems rarely integrate reliable sensor networks and data processing algorithms, making it impossible to monitor and automatically calibrate system health status (such as matrix humidity and equipment performance degradation) in real time, which affects the accuracy and reliability of management.

[0007] While some patented technologies have proposed modular containers or pallet designs based on recycled materials to address the aforementioned issues, there is still room for improvement in achieving interlocking structures for rapid and stable installation, as well as in deep integration with intelligent monitoring and self-calibration algorithms.

[0008] Therefore, there is an urgent need for a green roof system solution that integrates rapid deployment, intelligent monitoring, and automatic calibration functions to address the problems of complex construction, unreliable module connections, and lack of intelligent monitoring and self-calibration capabilities in existing green roof technologies. Summary of the Invention

[0009] The purpose of this invention is to provide a snap-on rapid deployment device and self-calibration method for green roofs, so as to solve the technical problems mentioned in the background art.

[0010] To achieve the above objectives, the present invention provides the following technical solution: a snap-on rapid deployment device for green roofs, comprising multiple modular tray units, a snap-on interlocking system, a substrate condition monitoring system, and a central controller;

[0011] The modular tray unit is used for planting green plants; such as drought-resistant growth medium containing highly absorbent polymers or ecological porous fiber cotton, for planting Sedum plants (such as Sedum lineare) and other green plants.

[0012] The modular tray unit has a downwardly extending protrusion at the bottom, and its hollow interior forms a water storage space.

[0013] The snap-lock system is placed around the modular pallet unit and is used to assemble and fix multiple adjacent modular pallet units.

[0014] The substrate condition monitoring system is used to monitor the humidity parameters of the substrate;

[0015] The central controller is connected in communication with the substrate condition monitoring system and is used to control the irrigation equipment based on the monitoring data.

[0016] Furthermore, the modular tray unit includes a tray body, a lightweight cultivation substrate, and a porous non-woven fabric cover;

[0017] The tray body is a waterproof bottom layer made of recycled thermoplastic material. The bottom of the tray body is provided with a downwardly extending protrusion. The interior of the protrusion is hollow, and the hollow part of the protrusion is used to form a water storage space.

[0018] The lightweight cultivation substrate is made of highly absorbent polymers, lightweight porous aggregates, fibrous materials, and organic / inorganic cultivation materials. The lightweight cultivation substrate is used to form a drought-resistant growth culture medium.

[0019] The top of the lightweight cultivation substrate is covered with a porous non-woven fabric cover.

[0020] Furthermore, the snap-lock system includes a snap and a groove. The side wall of the pallet body is provided with a snap and a groove. The groove matches the snap, and the snap can be snapped into the groove of an adjacent pallet body.

[0021] Furthermore, the groove can be selectively filled with adhesive depending on whether separation is required. When multiple pallet bodies need to be separated, no adhesive is placed in the groove; when multiple pallet bodies do not need to be separated, adhesive is placed in the groove.

[0022] Furthermore, the bottom of the tray body is provided with multiple absorbent cotton strips, the bottom of which is located in the protrusion, and the top of which is inserted into the lightweight cultivation substrate. The absorbent cotton strips are used to transfer the water stored in the protrusion to the lightweight cultivation substrate. The protrusion can store excess rainwater during rainfall, forming additional water storage space.

[0023] Furthermore, the substrate condition monitoring system includes a moisture sensor distributed within the lightweight cultivation substrate, a high-precision reference sensor located in an unobstructed position in the center of the roof, and area sensors for each modular tray unit. The reference sensor includes at least a light sensor.

[0024] Furthermore, the central controller employs a self-calibration procedure, the execution of which includes at least the following steps:

[0025] Illumination data is acquired using a reference sensor and a region sensor;

[0026] Based on reference sensor readings, sunrise and sunset events are adaptively detected, and a sliding calibration window is dynamically defined.

[0027] Within the sliding calibration window, the readings of the regional sensor are smoothed and outliers are removed;

[0028] Calculate the calibration factor for the sensors in each region and correct the sensor readings in real time;

[0029] Analyze the historical trend of the calibration factor to predict the initial factor for the next day.

[0030] Furthermore, the dynamically defined sliding calibration window includes calculating the length of day, dividing the actual length of day into N equal parts to determine the half-width of the sliding window, thereby allowing the window width to be dynamically adjusted with the seasons;

[0031] The smoothing process employs a five-point median filtering method, and outlier removal is based on the median of the filtered data sequence and a preset absolute deviation range.

[0032] A self-calibration method for green roofs includes at least the following steps:

[0033] S1: Data acquisition, environmental data is acquired through reference sensors and area sensors at a preset sampling frequency;

[0034] S2: Adaptive calibration, based on the reference sensor data of S1, identifies environmental inflection points, dynamically determines the calibration window, and calibrates the regional sensor data;

[0035] S3: Intelligent control, using calibrated data to regulate irrigation equipment or generate status reports through a central controller;

[0036] S4: Trend prediction. Based on the historical data of calibration factors over the past few days, calculate the median change and generate the initial predicted value of the calibration factor for the next day.

[0037] Furthermore, the adaptive calibration also includes a neighborhood consistency check;

[0038] The neighborhood consistency check includes at least the following steps:

[0039] If the difference between the calibration factor of a sensor in a certain area and the calibration factor of a sensor in a neighboring area exceeds a threshold, the median value of the calibration factor in the neighboring area will be used for replacement.

[0040] Furthermore, the trend prediction involves performing a first-order difference on the most recent K-day calibration factor sequence and taking the median of the difference sequence as the change, thereby generating the initial calibration factor for the next day, where K is an integer from 3 to 14, preferably 7.

[0041] Compared with the prior art, the beneficial effects of the present invention are:

[0042] 1. This invention enables rapid deployment and reduces load. Its modular design and snap-fit ​​interlocking structure make installation as simple as assembling building blocks, significantly shortening the construction cycle and reducing reliance on specialized skills. The lightweight tray units and cultivation substrate effectively reduce the load on the roof structure.

[0043] 2. This invention enables intelligent and precise water management. Through an integrated sensor network and self-calibration algorithm, the system can perceive the real moisture status of the substrate in real time and drive the irrigation system to make precise water replenishment, avoiding water waste or insufficient water supply caused by traditional fixed-time irrigation, and improving water-saving efficiency and plant survival rate.

[0044] 3. The invention offers high reliability and maintainability, with the snap-fit ​​connection providing excellent mechanical stability. Functions such as neighborhood consistency checking in the self-calibration method can automatically identify sensor anomalies, ensuring the long-term reliability of the system. The modular design also makes localized repairs or replacements exceptionally simple, requiring only the release of the snap-fit ​​in the specific area.

[0045] 4. This invention is adaptive and environmentally friendly. The self-calibration method can adapt to changes in different seasons and weather conditions without the need for manual parameter adjustments. Furthermore, the equipment can be manufactured using recycled materials (such as thermoplastics and superabsorbent polymers recovered from waste diapers), embodying the concept of a circular economy and demonstrating significant environmental benefits. Attached Figure Description

[0046] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0047] Figure 1 This is a cross-sectional view of the modular tray unit of the present invention;

[0048] Figure 2 This is a partial enlarged view of the connection between the two modular tray units of the present invention;

[0049] Figure 3 This is a flowchart of the self-calibration method of the present invention;

[0050] Figure 4 This is a schematic diagram of the roof system deployment of the present invention.

[0051] In the diagram: 1. Tray body; 2. Protrusion; 3. Buckle; 4. Groove; 5. Lightweight cultivation substrate; 6. Porous non-woven fabric cover; 7. Absorbent cotton strip; 8. Adhesive; 9. Reference sensor; 10. Area sensor. Detailed Implementation

[0052] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments.

[0053] The device of this invention includes a modular tray unit, a snap-locking system, a substrate condition monitoring system, and a central controller. The modular tray unit features a lightweight design, is filled with a lightweight cultivation substrate, and has a multi-functional protruding structure at the bottom. The snap-locking system achieves rapid positioning and secure connection through snap-locking and groove designs around the module. The self-calibration method is based on a network of reference sensors and regional sensors deployed on the roof. It determines the calibration window through an adaptive detection algorithm, employs multi-level filtering and neighborhood consistency checks, and calculates the calibration factor in real time, achieving intelligent monitoring and precise control of the green roof system. This patented technology solves the problems of complex construction, difficult maintenance, and lack of intelligent control in traditional green roofs, enabling rapid deployment and intelligent operation and maintenance of green roofs.

[0054] Example 1:

[0055] Please see Figure 1 , Figure 2 and Figure 4 A snap-on rapid deployment device for green roofs includes multiple modular tray units, a snap-on interlocking system, a substrate condition monitoring system, and a central controller;

[0056] Modular tray units are used for planting green plants;

[0057] The modular tray unit includes a tray body 1, a lightweight cultivation substrate 5, and a porous non-woven fabric cover 6;

[0058] The tray body 1 is a waterproof bottom layer made of recycled thermoplastic material. The bottom of the tray body 1 is provided with a downwardly extending protrusion 2. The interior of the protrusion 2 is hollow, and the hollow part of the protrusion 2 is used to form a water storage space.

[0059] The lightweight cultivation substrate 5 is composed of highly absorbent polymers, lightweight porous aggregates, fibrous materials, and organic / inorganic cultivation components. The lightweight cultivation substrate 5 is used to form a drought-resistant growth medium.

[0060] The top of the lightweight cultivation substrate 5 is covered with a porous non-woven fabric cover 6.

[0061] The bottom of the tray body 1 is provided with multiple absorbent cotton strips 7. The bottom of the absorbent cotton strips 7 is located in the protrusion 2, and the top of the absorbent cotton strips 7 is inserted into the lightweight cultivation substrate 5. The absorbent cotton strips 7 are used to transfer the water stored in the protrusion 2 to the lightweight cultivation substrate 5. The protrusion 2 can store excess rainwater during rainfall, forming an additional water storage space.

[0062] Modular tray units are the basic building blocks of the system. They are filled with lightweight cultivation substrates (such as drought-resistant growth media containing highly absorbent polymers or eco-friendly porous fiber cotton) for growing Sedum plants (such as Sedum lineare) and other greenery. The bottom of the tray features downward-extending, hollow sections that store excess rainwater during rainfall, creating additional water storage space.

[0063] The snap-lock system is placed around the perimeter of the modular pallet unit and is used to assemble and fix multiple adjacent modular pallet units.

[0064] The snap-lock system includes a snap-lock 3 and a groove 4. The side wall of the pallet body 1 is provided with a snap-lock 3 and a groove 4. The groove 4 matches the snap-lock 3, and the snap-lock 3 can be snapped into the groove 4 of the adjacent pallet body 1.

[0065] Adhesive 8 can be selectively provided in the groove 4 depending on whether separation is required. When multiple pallet bodies 1 need to be separated, adhesive 8 is not provided in the groove 4. When multiple pallet bodies 1 do not need to be separated, adhesive 8 is provided in the groove 4.

[0066] The snap-fit ​​interlocking system is key to rapid deployment. Modular pallet units feature snap-fits (such as tongues) and matching grooves around their perimeter. During installation, simply align the snap-fit ​​of one module with the groove of the adjacent module to achieve quick positioning and mechanical interlocking. To further enhance connection stability, adhesive can be applied inside the grooves to create a permanent connection between adjacent pallets. This design eliminates complex tools and construction processes, significantly improving installation efficiency.

[0067] The substrate condition monitoring system is used to monitor the humidity parameters of the substrate;

[0068] The substrate condition monitoring system includes a moisture sensor distributed within the lightweight cultivation substrate 5, a high-precision reference sensor 9 located in the unobstructed central position of the roof, and area sensors 10 for each modular tray unit. The reference sensor 9 includes at least a light sensor. The substrate condition monitoring system forms an all-round sensing network through multiple sensors, and collects key parameters such as substrate humidity and ambient light in real time to provide data support for intelligent regulation.

[0069] The central controller employs a self-calibration procedure, which includes at least the following steps:

[0070] Illumination data is acquired using reference sensor 9 and area sensor 10;

[0071] Based on the readings of reference sensor 9, sunrise and sunset events are adaptively detected, and a sliding calibration window is dynamically defined;

[0072] Within the sliding calibration window, the readings of the area sensor 10 are smoothed and outliers are removed;

[0073] Calculate the calibration factor of sensor 10 in each region and correct the sensor readings in real time;

[0074] Analyze the historical trend of the calibration factor to predict the initial factor for the next day.

[0075] The central controller communicates with the substrate condition monitoring system to control irrigation equipment based on monitoring data.

[0076] The dynamic definition of the sliding calibration window involves calculating the length of day, dividing the actual length of day into N equal parts (e.g., 20 equal parts) to determine the half-width of the sliding window, so that the window width is dynamically adjusted with the seasons;

[0077] The smoothing process employs a five-point median filtering method, with outlier removal based on the median of the filtered data sequence and a preset absolute deviation range.

[0078] The central controller, acting as the system's "brain," communicates with the substrate condition monitoring system and receives sensor data. Internally, the controller runs self-calibration algorithms and intelligent decision-making models, precisely controlling irrigation equipment and other actuators based on the processed data to achieve on-demand irrigation and intelligent monitoring of the system's status.

[0079] The equipment operates as follows:

[0080] Preparation: Clean the roof base layer and ensure the waterproofing layer is intact. Calculate the required number of modular tray units based on the area and shape of the roof to be greened.

[0081] Laying the starting module: Begin by placing the first modular tray unit at one corner of the roof. Ensure accurate positioning and good contact between the bottom protrusion 2 and the roof surface.

[0082] Snap-on connection: Align the snap-on 3 on the side of the second modular tray unit with the groove 4 on the side of the first unit, and gently press down vertically until you hear a "click," indicating that the connection is in place. Repeat this process to connect subsequent modules in sequence until the entire predetermined area is covered. For areas requiring particularly secure connections, inject an appropriate amount of waterproof adhesive 8 into the groove 4 before snapping it in.

[0083] Sensor Placement and Wiring: Install the high-precision reference sensor 9 in an unobstructed location at the center of the roof, with its panel facing horizontally upwards. Embed the area sensors 10 in the cultivation substrate of the selected modular tray units. Connect all sensors to the corresponding interfaces on the central controller according to the instruction manual.

[0084] Planting: Plant pre-acclimatized Sedum seedlings (such as Sedum lineare) or other plants suitable for rooftop growth in the cutting holes or planting areas of the modular tray unit.

[0085] System power-on and commissioning: Connect the central controller power supply, configure network parameters (if necessary, connect to an NTP server for time calibration), and start the system. Set irrigation parameter thresholds in the controller interface and activate the self-calibration program.

[0086] Example 2:

[0087] See Figure 3 A self-calibration method for green roofs includes at least the following steps:

[0088] S1: Data acquisition, environmental data is acquired at a preset sampling frequency through reference sensor 9 and area sensor 10;

[0089] The roof reference sensor and the area sensors 10 synchronously collect environmental data (such as light intensity) at preset time intervals (e.g., every minute). The sensor clocks can be synchronized with the server via Network Time Protocol (NTP) to ensure time consistency.

[0090] S2: Adaptive calibration, based on the reference sensor data of S1, identifies environmental inflection points, dynamically determines the calibration window, and calibrates the data of regional sensor 10;

[0091] First, by performing gradient analysis on the readings of the roof reference sensor throughout the day (e.g., calculating the central differential gradient), the inflection points of environmental changes such as sunrise and sunset are adaptively detected. Then, based on the calculated actual length of day, the width and position of the sliding calibration window are dynamically determined (e.g., dividing the length of day into 20 equal parts as half the window width), so that the window can automatically adjust with seasonal changes and always cover the dawn and dusk periods when light changes are most drastic.

[0092] Calibration factor calculation and real-time correction: Using the retained valid data points, the calibration factor (e.g., the mean of the ratio sequence) of sensor 10 in each region is calculated. This factor is used to correct the readings of sensor 10 in that region throughout the day. Simultaneously, a neighborhood consistency check mechanism is introduced: if the calibration factor of a sensor differs significantly from the median calibration factor of its neighboring sensors, the sensor is deemed to be malfunctioning, and the neighborhood median is automatically used as a substitute to prevent single-point failures from affecting the overall reliability of the system.

[0093] S3: Intelligent control, using calibrated data to regulate irrigation equipment or generate status reports through a central controller;

[0094] S4: Trend prediction. Based on the historical data of calibration factors over the most recent few days (e.g., 7 days), calculate the median change and generate the initial predicted value of the calibration factor for the next day.

[0095] The calculated calibration factors are stored in the historical database daily. By analyzing the trend of factor changes over a recent period (e.g., 7 days) (such as the median of the first difference), the initial calibration factor for the next day is predicted. This helps the system converge quickly during daily startup and reduces cold start time.

[0096] Based on Example 2, a specific application is proposed:

[0097] Assume the system is running on a sunny summer day:

[0098] Data acquisition: The reference sensor (range 0-200000 lux) and all area sensors acquire data once per minute and store it in the central controller database.

[0099] Sunrise detection and window determination: The central controller calculates the sunrise threshold using the median of the previous day's total illumination gradient plus the median absolute deviation (MAD); it calculates the gradient using reference sensor data from 4:00 AM to 10:00 AM of the current day, and determines the sunrise time T_sunrise when the gradient first exceeds the threshold; T_day is determined by dividing the measured day length of the previous day by the astronomical calendar day length; the calibration window is set to... [T_sunrise+5min,T_sunrise+T_day / 20] (Only the area after sunrise is retained to avoid low-light environments).

[0100] Calibration calculation: Within the calibration window, calculate the ratio of the readings of the regional sensor S1 to the reference sensor (S1 reading / reference reading) sequence; after performing five-point median filtering and 3σ criterion outlier removal on the comparison value sequence, take the mean of the sequence as the calibration factor K1=1.05 for S1 (indicating that the S1 reading is 5% higher than the reference).

[0101] Real-time correction and neighborhood verification: The original reading V_raw of S1 on the same day is corrected according to the formula V_corrected=V_raw / K1; at the same time, the absolute difference between S1's K1 and the median of the calibration factors of the three surrounding sensors is verified. If it is ≤ the set threshold of 0.1, then K1=1.05 is accepted.

[0102] Trend prediction: The system records the effective K1 of S1 on the current day, combines it with the historical effective K1 data of the previous 6 days, calculates the median of the first difference over 7 days, and predicts the initial calibration factor for the next day as 1.04 according to K1_predicted=K1 on the current day+median of the difference.

[0103] In summary:

[0104] The snap-on rapid deployment device and self-calibration method for green roofs provided by this invention have a clear structure and feasible method. They can be widely applied to greening renovation projects on flat or gently sloping roofs of various new or existing buildings, including commercial buildings, industrial plants, and residential buildings. This effectively lowers the construction threshold and maintenance costs of green roofs, improves the level of intelligent management, and has positive significance for promoting the development of green buildings and improving the urban ecological environment. It possesses good industrial practicality and promotional value.

[0105] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the invention can be implemented in other specific forms without departing from its spirit or essential characteristics. Therefore, the embodiments should be considered in all respects as exemplary and non-limiting, and the scope of the invention is defined by the appended claims rather than the foregoing description. Thus, all variations falling within the meaning and scope of equivalents of the claims are intended to be included within the present invention. No reference numerals in the claims should be construed as limiting the scope of the claims.

Claims

1. A snap-on rapid deployment device for green roofs, characterized in that: +0- Includes multiple modular tray units, a snap-lock interlock system, a matrix condition monitoring system, and a central controller; The modular tray unit is used for planting green plants; the bottom of the modular tray unit is provided with a downwardly extending protrusion, and the interior is hollow to form a water storage space; The snap-lock system is placed around the modular pallet unit and is used to assemble and fix multiple adjacent modular pallet units. The substrate condition monitoring system is used to monitor the substrate's humidity parameters and roof environmental parameters; The central controller is connected in communication with the substrate condition monitoring system and is used to control the irrigation equipment based on the monitoring data.

2. The snap-on rapid deployment device for green roofs according to claim 1, characterized in that: The modular tray unit includes a tray body (1), a lightweight cultivation substrate (5), and a porous non-woven fabric cover (6). The tray body (1) is a waterproof bottom layer made of recycled thermoplastic material. The bottom of the tray body (1) is provided with a downwardly extending protrusion (2). The interior of the protrusion (2) is hollow, and the hollow part of the protrusion (2) is used to form a water storage space. The lightweight cultivation substrate (5) is made of highly absorbent polymers, lightweight porous aggregates, fiber materials and organic / inorganic cultivation materials. The lightweight cultivation substrate (5) is used to form a drought-resistant growth culture medium. The top of the lightweight cultivation substrate (5) is covered with a porous nonwoven fabric cover (6).

3. The snap-on rapid deployment device for green roofs according to claim 2, characterized in that: The snap-lock system includes a snap (3) and a groove (4). The side wall of the pallet body (1) is provided with a snap (3) and a groove (4). The groove (4) matches the snap (3). The snap (3) can be snapped into the groove (4) of the adjacent pallet body (1).

4. The snap-on rapid deployment device for green roofs according to claim 3, characterized in that: The groove (4) can be selectively provided with adhesive (8) depending on whether separation is required. When multiple pallet bodies (1) need to be separated, adhesive (8) is not provided in the groove (4). When multiple pallet bodies (1) do not need to be separated, adhesive (8) is provided in the groove (4).

5. The snap-on rapid deployment device for green roofs according to claim 2, characterized in that: The bottom of the tray body (1) is provided with a plurality of absorbent cotton strips (7), the bottom of the absorbent cotton strips (7) is located in the protrusion (2), and the top of the absorbent cotton strips (7) is inserted into the lightweight cultivation substrate (5). The absorbent cotton strips (7) are used to transfer the water stored in the protrusion (2) to the lightweight cultivation substrate (5).

6. The snap-on rapid deployment device for green roofs according to claim 1, characterized in that: The substrate condition monitoring system includes a moisture sensor distributed in the lightweight cultivation substrate (5), a high-precision reference sensor (9) in the unobstructed central position of the roof, and a regional sensor (10) for each modular tray unit. The reference sensor (9) includes at least a light sensor.

7. The snap-on rapid deployment device for green roofs according to claim 6, characterized in that: The central controller employs a self-calibration program, the execution of which includes at least the following steps: Illumination data were collected using a reference sensor (9) and a region sensor (10); Based on the readings of the reference sensor (9), sunrise and sunset events are adaptively detected, and a sliding calibration window is dynamically defined; Within the sliding calibration window, the readings of the area sensor (10) are smoothed and outliers are removed; Calculate the calibration factor of the sensor (10) in each region and correct the sensor readings in real time; Analyze the historical trend of the calibration factor to predict the initial factor for the next day.

8. The snap-on rapid deployment device for green roofs according to claim 7, characterized in that: The dynamically defined sliding calibration window includes calculating the length of day, dividing the actual length of day into N equal parts to determine the half-width of the sliding window, so that the window width is dynamically adjusted with the seasons. The smoothing process employs a five-point median filtering method, and outlier removal is based on the median of the filtered data sequence and a preset absolute deviation range.

9. A self-calibration method for green roofs, used in the snap-on rapid deployment device for green roofs as described in any one of claims 1-8, characterized in that: At least the following steps are included: S1: Data acquisition, environmental data is acquired at a preset sampling frequency through reference sensor (9) and area sensor (10); S2: Adaptive calibration, based on the reference sensor data of S1, identifies environmental inflection points, dynamically determines the calibration window, and calibrates the regional sensor data; S3: Intelligent control, using calibrated data to regulate irrigation equipment or generate status reports through a central controller; S4: Trend prediction. Based on the historical data of calibration factors over the past few days, calculate the median change and generate the initial predicted value of the calibration factor for the next day.

10. A self-calibration method for a green roof according to claim 9, characterized in that: The adaptive calibration also includes a neighborhood consistency check; The neighborhood consistency check includes at least the following steps: When the difference between the calibration factor of a sensor (10) in a certain area and the calibration factor of the sensor (10) in the adjacent area exceeds a threshold, the median value of the calibration factor in the neighboring area is used for replacement.

Images