A residential building carbon emission monitoring system and an accounting method thereof

Abstract

The application provides a residential building carbon emission monitoring system and an accounting method thereof, to realize self-cleaning of the structure and reduce the frequency of manual maintenance. In the accounting method of the residential building carbon emission monitoring system, a shell, a lifting mechanism, a monitoring mechanism, a power mechanism, a dustproof mechanism and a control mechanism are arranged; air around the residential building is contacted with the monitoring mechanism through the shell; the monitoring mechanism collects data of the current surrounding air; then the monitoring mechanism transmits the monitoring value to the control mechanism and stores the monitoring value in the control mechanism; data fusion and feature inversion are performed by using a machine learning algorithm and a plurality of satellite monitoring data; original data are corrected by a deep learning algorithm and ground monitoring data; finally, data inversion is performed on the monitoring area until detailed carbon emission data of a small-scale area are obtained.

Description

Technical Field

[0001] This invention relates to the field of carbon emission monitoring technology, specifically to a residential building carbon emission monitoring system and its calculation method. Background Technology

[0002] The residential building carbon emission monitoring system monitors the carbon emissions around residential buildings to determine the carbon emissions of daily life in different residential buildings, and takes corresponding measures to reduce carbon emissions based on the data, thereby reducing the heat island effect.

[0003] Current carbon emission monitoring equipment mostly involves setting up multiple devices at different locations within a community to obtain an average data value for the entire monitoring area. Regarding measurement, carbon dioxide has a molecular weight of 44 g / mol, while air has a molecular weight of approximately 28.8 g / mol. This significant difference means that carbon dioxide takes longer to collide and diffuse with other air molecules, resulting in a slower diffusion rate. Furthermore, carbon dioxide is denser than air, meaning it falls to the ground and forms concentration gradients in low-lying areas. Additionally, carbon dioxide levels are generally lower in the morning due to solar energy promoting photosynthesis, and higher at night when photosynthesis ceases. These factors contribute to the inaccuracy of carbon dioxide monitoring equipment.

[0004] Under the influence of various factors and changes in carbon dioxide, Chinese invention patent CN202210592748.X (publication date: 2022-08-09) discloses a carbon emission monitoring device based on environmental protection. Its structure includes a frame and a monitoring device. A walking device is installed below the frame, and a lifting device and a positioning device are installed above the frame. A fixed plate is installed above the lifting device, and the monitoring device is mounted on the fixed plate. The positioning device is used to fix the frame. By employing the lifting device and positioning device, the monitoring device is driven to a certain height to monitor carbon emissions, and the frame is positioned by the positioning device to improve the accuracy of carbon emission monitoring.

[0005] Based on the differences in carbon dioxide levels between morning and evening, and the varying levels of human activity throughout the day and night, the use of this device presents challenges. Without adapting the circuit signal structure to the current conditions, the device would become bulky, and its relocation would require consideration of various factors, including whether it would disrupt traffic or create obstacles. Therefore, the device's effectiveness in monitoring carbon emissions has not yet fully and reasonably resolved these issues.

[0006] To address these issues and improve the performance of carbon emission monitoring equipment to obtain more accurate data, this invention provides a residential building carbon emission monitoring system and its calculation method to solve the aforementioned problems. Summary of the Invention

[0007] In view of the deficiencies of the prior art, the purpose of this invention is to provide a residential building carbon emission monitoring system and its calculation method to solve the problems that: the angle of the monitoring equipment requires additional cost input for adjustment and cannot be effectively adjusted according to the distribution of carbon dioxide.

[0008] To solve the above-mentioned technical problems, the technical solution adopted by the present invention is as follows: a carbon emission monitoring system for residential buildings is provided, including a shell, a lifting mechanism, a monitoring mechanism, a power mechanism, a dust prevention mechanism, and a control mechanism;

[0009] The housing is a rectangular box, and the lifting mechanism is detachably connected to the upper surface of the housing. The lifting mechanism has a spiral blade structure. The monitoring mechanism is composed of multiple monitors and is detachably connected to the lifting mechanism. The power mechanism is fixedly connected to the lower surface of the lifting mechanism. The dustproof mechanism is arranged around the lifting mechanism, and the monitoring mechanism is located inside the dustproof mechanism. Both the monitoring mechanism and the power mechanism are electrically connected to the control mechanism.

[0010] Preferably, the upper surface of the housing is provided with a cover, the cover and the housing are detachably connected, the cover has an opening at the central axis position, the inner edge of the opening is provided with a guide groove, and the diameter of the opening is smaller than the side length of the cover; the edge of the housing is provided with a ventilation hole, and the ventilation hole on one side of the housing is preferably a rectangular through hole, and the ventilation hole is provided with an inclined angle, which is set between 15-30°.

[0011] Preferably, a nylon brush is fixedly connected to the lower surface of the cover; the length of the nylon brush bristles is within the range of 20-25mm, the diameter of the nylon brush bristles is within the range of 0.3-0.5mm, and the distance between the nylon brush bristles is between 0.5-1mm. The nylon brush can directly contact the dustproof mechanism by rotating the lifting mechanism; a warning structure is provided on the edge of the housing, which can be an LED, halogen lamp, fluorescent lamp, etc., preferably an LED with a brightness of more than 1000cd / ㎡, and the color temperature is set between 5000K and 6500K depending on the occasion. The warning structure is electrically connected to the control mechanism.

[0012] Preferably, the lifting mechanism includes a spiral blade and a snap-fit ​​plate; one end of the spiral blade is fixedly connected to the lower surface of the snap-fit ​​plate, the snap-fit ​​plate has a conical structure, and the lower surface of the snap-fit ​​plate has a recessed structure, the outer surface of the snap-fit ​​plate has the same size as the opening; the interior of the spiral blade has a hollow structure, the size of which is larger than the maximum diameter of the motion detector; the interior of the spiral blade is provided with multiple spiral structures.

[0013] Preferably, the lower end of the spiral blade is provided with a closed-loop blade, so that the lifting mechanism rotates at the limit height; the part of the outer surface of the spiral blade that contacts the guide groove has a friction force much smaller than its own weight.

[0014] Preferably, the monitoring mechanism includes a mobile monitor and a fixed monitor; the mobile monitor is snapped into the spiral blade, preferably at the bottom of the lifting mechanism; multiple fixed monitors are provided, evenly distributed along the upper surface of the spiral blade, preferably four fixed monitors per revolution; the mobile monitor is electrically connected to the control mechanism, preferably a flying monitor, and equipped with a camera function; the fixed monitor uses infrared optics. sensor.

[0015] Preferably, the power mechanism includes a solar panel, an inverter, a battery, a motor, and a connecting shaft; the solar panel is fixedly connected to the upper surface of the cover and the snap-fit ​​plate, the solar panel is electrically connected to the battery, the battery is electrically connected to the inverter, and the motor is electrically connected to the battery and the control mechanism; the connecting shaft is fixedly connected to the motor and to the lower end of the spiral blade.

[0016] Preferably, the dustproof mechanism includes a frame and a mesh screen; the upper end of the frame is fixedly connected to the snap-fit ​​plate, the mesh screen is arranged along the frame, the upper part of the frame is not provided with the mesh screen, the exposed part of the frame is not less than twice the height of the mobile monitor, the lower end of the dustproof mechanism covers the fixed monitoring mechanism, the mesh size of the mesh screen is between 0.5 and 5 mm, the density of the mesh screen should be between 100 g / m² and 200 g / m², the thickness is between 0.5 mm and 1.5 mm, and the material of the mesh screen is selected to be wear-resistant, corrosion-resistant, and UV-resistant, preferably polypropylene, and the color is preferably light-colored to reduce heat accumulation.

[0017] This invention also provides a method for calculating carbon emission monitoring data of residential buildings, comprising the following steps:

[0018] S1: When the power mechanism is not in operation, the monitoring unit is located inside the housing, and the air around the building comes into contact with the fixed monitor. When the power mechanism rotates, the spiral blades rise from inside the housing under the action of the guide groove. Subsequently, the mobile monitor leaves the inside of the spiral blades, and the fixed monitor transmits the signal via infrared optics. Sensors collect data from surrounding emission sources, and mobile monitors use infrared optics. The sensor collects multi-view building data and transmits the monitored values ​​to the host computer database via RS485 protocol. Then, a Gaussian inversion model is established through machine learning algorithm to make a preliminary estimate of carbon emissions in the monitored area and establish a dataset.

[0019] S2: Use the LabelImg tool to label target objects such as buildings in the dataset, including drawing bounding boxes on the image, identifying the location and size of the target objects, and then save the labeling results to the corresponding annotation file to provide useful training data for subsequent machine learning algorithms;

[0020] S3: Weave different satellite monitoring data using methods such as weighted averaging or principal component analysis to obtain the best fusion effect; then use vector machines and neural networks to perform feature inversion, extract and analyze features from the fused data, and invert some features of the target object; finally, use labeled datasets to test and optimize the algorithm to obtain the best inversion effect.

[0021] S4: Use convolutional neural networks or recurrent neural networks for interpolation. Train a deep learning model by taking the valid data around the missing value as input, predict the value of the missing value, and complete the information. Then, the data from the monitoring agency can be used to adjust the meteorological data monitored by satellite.

[0022] S5: The inverse distance weighted interpolation method is used to perform data inversion on the detection results of the multi-view full-dimensional dynamic convolution YOLO-ODConv building detection algorithm, and the carbon emissions of the building carbon emission submerged (static) and operational (active) phases are integrated to obtain detailed carbon emission data for small-scale areas; among which, the carbon emissions of the building submerged (static) phase are used for identification, including carbon emissions of building materials, carbon emissions of the construction process, and carbon emissions of material transportation; the carbon emissions of the building operation (activity) phase, that is, during the operation period, are divided into regional building base carbon emissions, building incremental carbon emissions, and the amount of carbon emission reduction from renewable energy.

[0023] The beneficial effects of this invention are as follows:

[0024] 1. By combining the lifting mechanism structure with the housing, the monitor can automatically rise and fall according to the intensity of sunlight, and can perform multi-angle monitoring during the rising and falling process, thereby obtaining more comprehensive monitoring results. At the same time, based on the specific monitoring effect, the data can be made more accurate by using the set detection mechanism and the calculation method provided by the present invention.

[0025] 2. The present invention, through the setting of lifting mechanism and dustproof mechanism, enables the monitoring mechanism to monitor from multiple angles while avoiding the impact of the environment on the equipment when there is a large flow of people. In addition, the equipment can achieve self-cleaning of the structure set inside the shell through each lifting and lowering, reducing the frequency of manual maintenance.

[0026] 3. This invention combines static and dynamic monitoring mechanisms. By using a lifting and power mechanism, the fixed monitoring equipment can be deployed to monitor a wider range of angles when there are many factors affecting carbon dioxide. When humidity is high or the change in carbon dioxide is small, it can enter a fixed monitoring mode to obtain more accurate carbon dioxide content data and achieve continuous monitoring, thus better understanding the trend of carbon dioxide changes.

[0027] 4. This invention addresses the challenges of data gaps and small-scale satellite monitoring by combining multiple monitoring data sources and machine learning algorithms. Deep learning algorithms provide more detailed and comprehensive carbon emission raster data for the monitored area, resulting in more accurate inversion results. Furthermore, the use of various machine learning and deep learning algorithms not only improves computational efficiency but also significantly reduces manual computation costs. The operation and calculation methods are simpler and easier to understand, and the calculation results are far superior to those from other single-carbon satellite monitoring inversion methods. Attached Figure Description

[0028] To more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the specific embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0029] Figure 1 This is an overall schematic diagram of the present invention;

[0030] Figure 2 This is an overall front view of the present invention;

[0031] Figure 3 This is an overall cross-sectional view of the present invention;

[0032] Figure 4This is a schematic diagram of the housing of the present invention;

[0033] Figure 5 This is a schematic diagram of the overall cutaway of the present invention;

[0034] Figure 6 This is a schematic diagram of the dustproof mechanism of the present invention;

[0035] Figure 7 This is a schematic diagram of the lifting mechanism of the present invention;

[0036] Figure 8 This is a schematic diagram of the spiral blade of the present invention;

[0037] Figure 9 This is a schematic diagram of the nylon brush of the present invention;

[0038] Figure 10 This is a schematic diagram of the capping of the present invention;

[0039] Figure 11 This is a schematic diagram of the snap-fit ​​plate of the present invention;

[0040] Figure 12 This is a schematic diagram of the snap-fit ​​plate of the present invention from another perspective;

[0041] Figure 13 This is a flowchart of the present invention;

[0042] Figure 14 This is a network structure diagram of the YOLO-ODConv algorithm of the present invention.

[0043] In the diagram: 1. Housing; 11. Cover; 12. Opening; 13. Guide groove; 14. Ventilation opening; 15. Nylon brush; 16. Warning structure; 2. Lifting mechanism; 21. Spiral blade; 22. Connecting plate; 3. Monitoring mechanism; 31. Movable monitor; 32. Fixed monitor; 4. Power mechanism; 41. Solar panel; 42. Inverter; 43. Battery; 44. Motor; 45. Connecting shaft; 5. Dustproof mechanism; 51. Frame; 52. Mesh screen. Detailed Implementation

[0044] The specific embodiments of the present invention will be further described below with reference to the accompanying drawings. It should be noted that these descriptions are for the purpose of aiding understanding the present invention and do not constitute a limitation thereof. Furthermore, the technical features involved in the various embodiments of the present invention described below can be combined with each other as long as they do not conflict with each other.

[0045] Example 1:

[0046] like Figure 1-3As shown, a carbon emission monitoring system for residential buildings is provided, including a housing 1, a lifting mechanism 2, a monitoring mechanism 3, a power mechanism 4, a dust prevention mechanism 5, and a control mechanism;

[0047] The housing 1 is a rectangular box, and the lifting mechanism 2 is detachably connected to the upper surface of the housing 1. The lifting mechanism 2 has a spiral blade structure. The monitoring mechanism 3 is composed of multiple monitors and is detachably connected to the lifting mechanism 2. The power mechanism 4 is fixedly connected to the lower surface of the lifting mechanism 2. The dustproof mechanism 5 is arranged around the lifting mechanism 2, and the monitoring mechanism 3 is located inside the dustproof mechanism 5. Both the monitoring mechanism 3 and the power mechanism 4 are electrically connected to the control mechanism.

[0048] The housing 1 is a hollow cylindrical box. A cover 11 is provided on the upper surface of the housing 1. The bottom surface of the cover 11 is rectangular. The four sides of the cover 11 extend upward and converge at a point. The cover 11 and the housing 1 are fixed by bolts and are detachable. An opening 12 is provided at the central axis of the cover 11. The opening 12 is the size of a vertically cut cylinder. A guide groove 13 is provided on the inner edge of the opening 12. The guide groove 13 is the same size as the width of the spiral blade 21. The diameter of the opening 12 is smaller than the side length of the cover 11. A vent 14 is provided on the edge of the housing 1. The vent 14 is a rectangular through hole on one side of the housing 1. The length of both ends of the vent 14 is smaller than the width of the housing 1. The vent 14 is inclined at an angle of 30°.

[0049] like Figure 4 As shown, the cover 11 is detachably connected to the housing 1. By opening the cover 11, maintenance and repair of the internal parts of the equipment can be achieved. The housing 1 is provided with an opening 12. The snap-fit ​​plate 22 can be snapped onto the cover 11 by snap-fit. This allows the internal structure of the housing 1 to be pushed out under the action of the opening 12. On the other hand, combined with the ventilation holes provided on the outer edge, the 30° angle ensures that rainwater is not easy to enter. At the same time, it allows the monitoring mechanism 3 inside the housing 1 to come into contact with the outside atmosphere.

[0050] A nylon brush 15 is fixedly connected to the lower surface of the cover 11. The bristles of the nylon brush 15 are 25mm long, 0.5mm in diameter, and 0.5mm apart. This effectively prevents dust and impurities from entering the cover 11, protecting the internal components of the machine. The nylon brush 15 can be directly contacted with the dustproof mechanism 5 by rotating the lifting mechanism 2, making it easy to clean and replace. A warning structure 16 is provided on the edge of the housing 1. The warning structure 16 uses an LED light with a brightness of 1000cd / ㎡ and a color temperature of 5000K, making the warning structure 16 easily identifiable at night.

[0051] like Figure 5-6 As shown, during the lifting mechanism 2's continued lifting and lowering process according to working conditions, since the nylon brush 15 is in close contact with the dustproof mechanism 5, and the dustproof mechanism 5 is exposed to the outside during the day, it is prone to accumulating a lot of dust, which can reduce its protective efficiency over time. Therefore, during the lifting process of the dustproof mechanism 5, the nylon brush 15 has a groove inside that matches the thread of the spiral blade 21, allowing the nylon brush 15 to directly contact the dustproof mechanism 5. The nylon brush 15 can clean the outer surface of the dustproof mechanism 5 from top to bottom. Due to the small diameter of the bristles and the uniform distance between the bristles, it can effectively remove impurities and dust from the mesh 52, keeping the mesh 52 clean. Furthermore, the uniform length of the bristles and the uniform distance between the bristles ensure the uniformity of contact between the bristles and the mesh 52 when brushing the mesh 52, avoiding the bristles being too dense or too sparse, thus ensuring that the bristles are evenly distributed across the entire surface of the mesh 52. Finally, the selected nylon structure gives the brush excellent wear resistance and corrosion resistance, making it suitable for different types of mesh 52.

[0052] like Figure 7-9 As shown, the lifting mechanism 2 includes a spiral blade 21 and a snap-fit ​​plate 22; the uppermost end of the spiral blade 21 is fixedly connected to the lower surface of the snap-fit ​​plate 22, the snap-fit ​​plate 22 has a conical structure, and the lower surface of the snap-fit ​​plate 22 has a recessed structure. The outer surface of the snap-fit ​​plate 22 is the same size as the opening 12; the interior of the spiral blade 21 has a hollow structure, and the size of the hollow structure is larger than the maximum diameter of the motion detector; the fixed monitor 32 is placed on the upper surface of the spiral blade 21; the motion monitor 31 is snapped onto the lowermost spiral blade 21; the lower end of the spiral blade 21 is a closed-loop blade; the part of the outer surface of the spiral blade 21 that contacts the guide groove 13 has a friction force much smaller than its own weight.

[0053] The monitoring mechanism 3 includes a mobile monitor 31 and a fixed monitor 32. The mobile monitor 31 is snapped onto the bottom end of the spiral blade 21. The fixed monitor 32 is configured with four units per revolution of the spiral blade 21, evenly distributed along the spiral blade 21. The mobile monitor 31 is electrically connected to the control mechanism and is configured as a flying monitor with a camera function. The fixed monitor 32 uses infrared optics. sensor.

[0054] The spiral blade 21 rotates under the action of the power mechanism 4. Due to the action of the limiting groove provided on the inner edge of the cover 11, the spiral blade 21 spirals upward along the limiting groove. The spiral blade 21 spirals upward, and at this time the snap plate 22 disengages from the cover 11.

[0055] The spiral blade 21 is spiral-shaped, allowing it to change angles as it rises. This enables the fixed monitor 32 inside to monitor from different heights and directions at the current position, resulting in more accurate data. The mobile monitor 31 is positioned at the bottom of the spiral blade 21 and can be engaged there. Because the spiral blade 21 has a hollow structure and the distance is sufficient for the mobile monitor 31 to move, it can move vertically from the top of the spiral blade 21 as it rises, achieving multi-view, multi-angle monitoring and identification. When the mobile monitor 31 is not needed, the fixed monitor 32 rises under the influence of the spiral blade 21, further enhancing the accuracy of the fixed monitoring mechanism 3 in environments with many influencing factors.

[0056] The power mechanism 4 includes a solar panel 41, an inverter 42, a battery 43, a motor 44, and a connecting shaft 45. The solar panel 41 is fixedly connected to the upper surface of the cover 11 and the snap-fit ​​plate 22. The solar panel 41 is electrically connected to the battery 43. The battery 43 is electrically connected to the inverter 42. The motor 44 is electrically connected to the battery 43 and the control mechanism. The connecting shaft 45 is fixedly connected to the motor 44 and to the lower end of the spiral blade 21.

[0057] like Figure 10-12As shown, the solar panel 41 is fixedly connected to and covers the upper surface of the cover 11 and the snap-fit ​​plate 22. During the day when the sunlight is strong, the solar panel 41 receives sunlight and converts the solar energy into electrical energy through the inverter 42. At this time, the power mechanism 4 starts to work, giving the lifting mechanism 2 a rotational force. The connecting shaft 45 is a telescopic structure, which gradually raises the spiral blade 21 under the action of the limiting groove. When the sunlight gradually falls, the solar panel 41 no longer receives solar energy, the power mechanism 4 stops working, and the spiral blade 21 loses its power source. Since the friction force is much less than the gravity, the spiral blade 21 gradually slides down under the action of gravity until it returns to the housing 1.

[0058] By setting up a light-sensing structure, the monitoring unit 3 can be extended during the day when carbon dioxide emission sources are more complex, so as to obtain a larger monitoring area; at night, when emission sources are relatively simple, the monitoring unit 3 can be retracted into the housing 1, which increases the safety of the equipment and adjusts the equipment to a low-energy consumption stage, making it more energy-efficient.

[0059] The dustproof mechanism 5 includes a frame 51 and a mesh 52. The upper end of the frame 51 is fixedly connected to the snap-fit ​​plate 22. The mesh 52 is arranged along the frame 51. The upper part of the frame 51 is not provided with the mesh 52. The exposed part of the frame 51 has a length not less than twice the height of the mobile monitor 31. The lower end of the dustproof mechanism 5 covers the fixed detection mechanism. The dustproof mechanism 5 is located 10mm inside the spiral blade 21. The mesh size of the mesh 52 is set to 2mm. The density of the mesh 52 is set to 100g / m². The thickness is set to 1mm. The material of the mesh 52 is a wear-resistant, corrosion-resistant, and UV-resistant material. In this embodiment, polypropylene is used. The color is also light-colored to reduce heat accumulation.

[0060] The dustproof mechanism 5 protects the monitoring mechanism 3 during the day and provides moisture protection at night. The dustproof mechanism 5, combined with the nylon brush 15, cleans the dust from the outer surface of the dustproof mechanism 5 during each ascent. Since the dustproof mechanism 5 does not cover the spiral blade 21, the spiral blade 21 can still slide normally within the limiting groove to complete the lifting process.

[0061] The overall working process is as follows: After the equipment is fully installed, as the sun rises, the solar panel 41 receives solar radiation, supplying energy to the motor 44. The motor 44 then begins to operate. Due to the telescopic function of the connecting shaft 45, the spiral blade 21 gradually rises under the action of the guide groove 13. The mesh 52 is cleaned during the ascent by the nylon brush 15, causing the fixed monitor 32 to be lifted. At the highest point, due to the closed-loop blades, the spiral blade 21 rotates, allowing the fixed monitor 32 to monitor from multiple directions and angles. After the sun sets, the sunlight gradually decreases, and the... When the motor 44 stops working, the lifting mechanism 2 loses its power source. Under the action of gravity, the lifting mechanism 2 overcomes the friction of the spiral blade 21 and begins to descend. The nylon brush 15 cleans the mesh 52 again until the lifting mechanism 2 returns to the bottom of the housing 1. For more precise monitoring, the mobile monitor 31 is used. The mobile monitor 31 is vertically separated from the housing 1 through the hollow structure on the spiral blade 21, thereby capturing multi-view data of the building to be monitored. Since the battery 43 and the inverter 42 are located inside the housing 1, the entire device can charge the mobile monitor 31 to ensure long-term use.

[0062] like Figure 13 As shown, firstly, the air around the residential building comes into contact with the monitoring agency 3 through the shell 1. The monitoring agency 3 collects data on the current surrounding air. Then, the monitoring agency 3 transmits the monitored values ​​and stores them in the control mechanism. Data fusion and feature inversion are performed using machine learning algorithms and data from multiple satellite monitoring systems. The original data is then corrected using deep learning algorithms and ground monitoring data. Finally, data inversion is performed on the monitoring area until detailed carbon emission data for a small-scale area is obtained.

[0063] This invention also provides a method for calculating carbon emission monitoring data of residential buildings, comprising the following steps:

[0064] S1: When the power mechanism 4 is not in operation, the monitoring unit 3 is located inside the housing 1, and the air around the building comes into contact with the fixed monitor 32; when the power mechanism 4 rotates, the spiral blade 21 rises from inside the housing 1 under the action of the guide groove 13, and then the mobile monitor 31 leaves from inside the spiral blade 21, passing through the fixed monitor 32 via infrared optics. The sensor collects data from surrounding emission sources, and the mobile monitor 31 uses infrared optical... Sensors collect multi-view building data, which is constructed by combining the four facades of the building. The monitored values ​​are then transmitted via RS485 protocol and stored in the host computer database. A Gaussian inversion model is then established using machine learning algorithms to make a preliminary estimate of carbon emissions in the monitored area and to build the dataset.

[0065] S2: The LabelImg tool was used to label target objects such as buildings in the dataset. This included drawing bounding boxes on the images, identifying the location and size of the target objects, and then saving the labeling results to the corresponding annotation files. This provides useful training data for subsequent machine learning algorithms. During model training, the error between the predicted and true values ​​is represented by a loss function. Backpropagation is used to continuously update the weight parameters to find the optimal solution; therefore, data labeling is necessary before the experiment to obtain true labels. The LabelImg tool was used to label all images in the VOC2007 dataset standard format, ultimately obtaining XML format label files.

[0066] S3: Weave different satellite monitoring data using methods such as weighted averaging or principal component analysis to obtain the best fusion effect; then use vector machines and neural networks to perform feature inversion, extract and analyze features from the fused data, and invert some features of the target object; finally, use labeled datasets to test and optimize the algorithm to obtain the best inversion effect.

[0067] A novel multi-view, full-dimensional dynamic convolutional YOLO-ODConv building detection algorithm is constructed. This algorithm is based on the YOLOv5 algorithm, fusing and embedding ODConv full-dimensional dynamic convolutions at layers 5, 7, and 10 of the backbone network to reduce computational parameters and improve algorithm accuracy. The formula for ODConv is:

[0068] ;

[0069] in, Represents the convolution kernel The attention scalar is the same as 1 in the equation; , and This represents three newly introduced attention points, each along the convolution kernel. Calculation of spatial dimension, input channel dimension, and output channel dimension; This represents multiplication operations along different dimensions of the kernel space. Here, and Use a multi-head attention module The calculations are based on this, which will be clarified later.

[0070] In ODConv, for convolution kernels (1) Different attention scalars are assigned to the convolution parameters (each filter) at k×k spatial locations; (2) For each convolution filter of Different attention scalars are assigned to the channels; (3) Assign different attention scalars Convolutional filter; (4) Assign an attention scalar to the entire convolution kernel. In principle, these four types of attention are complementary; multiply them progressively by the convolution kernel. The order of location, channel, filter, and kernel allows convolution operations to be performed at all spatial locations, across all input channels, and with all filters and kernels. All kernels are different, providing a performance guarantee for capturing rich contextual cues. Therefore, ODConv can significantly enhance the feature extraction capabilities of basic convolutional operations in CNNs. Furthermore, ODConv, using a single convolutional kernel, can compete with, and even outperform, the standard CondConv and DyConv, greatly reducing the additional parameters introduced to the final model. We provide extensive experiments to verify these advantages. Through the above formulas, we can clearly see that ODConv is a more general dynamic convolution. Moreover, when setting... and All components. and When all components are 1, only filtering attention is needed. ODConv will be simplified to: applying the input features to a convolutional filter under the condition of the SE variant, and then performing the convolution operation. Such an SE variant is a special case of ODConv.

[0071] like Figure 14As shown, due to the large number and complexity of the YOLO-ODConv algorithm network used in this stage, a large number of complex iterative calculations are required, so the experimental environment needs to meet relatively high requirements. The hardware combination is an Intel Core i9-12900K processor and an NVIDIA graphics card with RTX 3090 24 GB of video memory; in terms of the software environment, the experimental platform is based on the Windows 11 operating system and the deep learning framework PyTorch, utilizing CUDA 11.1 and CUDNN 8.0.5 for high-performance parallel computing; the dataset used by the YOLO-ODConv model is divided into training set, validation set and test set. The ratio of the training set to the test set is 7:3, and then 90% of the image data in the training set is used to input the network for training, and the remaining 10% of the image data is used for validation.

[0072] S4: Use convolutional neural networks or recurrent neural networks for interpolation. Train a deep learning model by taking the valid data around the missing value as input, predict the value of the missing value, and complete the information. Then, the data from the monitoring agency can be used to adjust the meteorological data monitored by satellite.

[0073] S5: The inverse distance weighted interpolation method is used to perform data inversion on the detection results of the multi-view full-dimensional dynamic convolution YOLO-ODConv building detection algorithm, and the carbon emissions of the building carbon emission submerged (static) and operational (active) phases are integrated to obtain detailed carbon emission data for small-scale areas; among which, the carbon emissions of the building submerged (static) phase are used for identification, including carbon emissions of building materials, carbon emissions of the construction process, and carbon emissions of material transportation; the carbon emissions of the building operation (activity) phase, that is, during the operation period, are divided into regional building base carbon emissions, building incremental carbon emissions, and the amount of carbon emission reduction from renewable energy.

[0074] At this stage, carbon emissions from both the stationary (submerged) and operational (active) phases of a building are measured together. The specific calculation formula is as follows:

[0075] ;

[0076] in, The carbon emissions from a building's static state are used for labeling, including carbon emissions from building materials, the construction process, and material transportation. It represents carbon emissions from building operations (activities), which during the operation period, and is divided into regional building base carbon emissions, carbon emissions from incremental building activities, and carbon emission reductions from renewable energy sources;

[0077] The formula for calculating the submerged (stationary) carbon emissions is as follows:

[0078] ;

[0079] in, This indicates the carbon emission coefficient for different types of building materials. Indicates the consumption of different types of building materials; This indicates that carbon emission coefficients for different types of construction processes are fitted using machine learning algorithms. This represents the building area of ​​different types of buildings detected based on the YOLO-ODConv algorithm; n represents the number of transport vehicles. Indicates the first Engine displacement of a car This indicates the carbon emission coefficient of various types of material transport vehicles. This indicates the mileage difference between different types of transport vehicles.

[0080] The formula for calculating the carbon emissions from the operation (activity) is as follows:

[0081] ;

[0082] ;

[0083] in, Indicates the increase in building carbon emissions. This indicates the reduction in carbon emissions from renewable energy buildings. Among them, express The first term represents the electricity resources available during the predetermined time period. A carbon emission coefficient, This indicates that the electricity consumption data corresponding to the first The amount of electricity consumed per carbon emission factor. Indicated in the pre-determined The water consumption during the specified time period This represents the carbon emission factor of water use. Indicates the number of processes processed within the predetermined time period. The weight of each type of waste Indicates the first The carbon emission coefficient of each type of waste. This refers to the reduction in carbon emissions from renewable energy sources, specifically the reduction in carbon emissions generated by renewable energy sources such as green spaces around buildings, photovoltaic systems, and solar energy systems. Indicates different amounts of renewable energy. This represents the carbon emission factor in the corresponding renewable energy conversion.

[0084] Example 2:

[0085] The mobile monitor 31 is located at the top of the spiral blade 21. A small motor is installed at the top of the spiral blade 21, which allows the mobile monitor 31 to rotate and capture images even when not in flight. At the same time, the number of fixed detectors inside the spiral blade 21 is increased, which can enhance the monitoring efficiency and improve the accuracy of the monitoring mechanism 3.

[0086] Those skilled in the art should understand that this invention is not limited to the above embodiments. The embodiments and descriptions in the specification are merely illustrative of the principles of the invention. Various changes, modifications, substitutions, and alterations can be made to these embodiments without departing from the spirit and scope of the invention. All such changes and modifications fall within the scope of the invention as claimed, which is defined by the appended claims and their equivalents. Even though the invention has been described with reference to specific exemplary embodiments, many different alternatives and modifications will become apparent to those skilled in the art. Furthermore, it should be noted that the circuit layout and identification / detection methods within the device of this invention can be omitted.

Claims

1. A method for monitoring and accounting for carbon emissions of a residential building, the method comprising: The resident building carbon emission monitoring and accounting method comprises the following steps: ​ S1: The monitoring mechanism (3) is located in the shell (1) in the non-working state of the power mechanism (4), and the air around the building contacts the fixed monitor (32); the power mechanism (4) rotates, the spiral blade (21) is lifted from the shell (1) under the action of the guide groove (13), and then the moving monitor (31) moves away from the hollow part of the spiral blade (21), and the fixed monitor (32) collects data from the surrounding emission sources through infrared optical sensors, and the moving monitor (31) collects multi-view building data through infrared optical sensors, and transmits the monitoring values to the host computer database through the RS485 protocol and stores them, and then establishes a Gaussian inversion model through a machine learning algorithm to preliminarily estimate the carbon emissions in the monitoring area and establish a data set. S2: Label the building target object in the data set by using the LabelImg tool, including drawing a bounding box on the image to identify the position and size of the target object, and then saving the labeling result to the corresponding annotation file to provide useful training data for subsequent machine learning algorithms; S3: Weighted average or principal component analysis method is used to fuse different satellite monitoring data to obtain the best fusion effect; then support vector machine and neural network are used for feature inversion to extract and analyze the features of the fused data, and some features of the target object are inverted; finally, the labeled data set is used for algorithm testing and optimization to obtain the best inversion effect; S4: Interpolation is performed using convolutional neural network or recurrent neural network, the effective data around the missing value is taken as input to train the deep learning model, the value of the missing value is predicted, and information completion is performed; then the data of the monitoring mechanism (3) is used to adjust the meteorological data of satellite monitoring; S5: The detection results of the multi-view full-dimensional dynamic convolution YOLO-ODConv building detection algorithm are inversed using the inverse distance weighted interpolation method, and the carbon emissions of the building in the construction and operation stages are fused to obtain detailed carbon emission data of small-scale areas.

2. A residential building carbon emissions monitoring system characterized by: A resident building carbon emission monitoring and accounting method for claim 1, the resident building carbon emission monitoring system comprises a shell (1), a lifting mechanism (2), a monitoring mechanism (3), a power mechanism (4), a dustproof mechanism (5) and a control mechanism; the shell (1) is a rectangular box, the lifting mechanism (2) is clamped on the upper surface of the shell (1), the lifting mechanism (2) is a spiral structure, the lifting mechanism (2) is used for adjusting the current height in cooperation with the shell (1) according to the current environmental conditions; the monitoring mechanism (3) is detachably connected with the lifting mechanism (2), the monitoring mechanism (3) is used to realize multi-angle conversion in cooperation with the lifting mechanism (2) to improve the monitoring efficiency; the power mechanism (4) is fixedly connected to the lower surface of the lifting mechanism (2), the power mechanism (4) is used to provide power to the lifting mechanism (2) and make strain according to the current environment; the dustproof mechanism (5) is clamped on the outer edge of the lifting mechanism (2), the dustproof mechanism (5) is used to protect the monitoring mechanism (3) while moving with the lifting mechanism (2); the monitoring mechanism (3) and the power mechanism (4) are electrically connected with the control mechanism.

3. A residential building carbon emissions monitoring system according to claim 2, wherein: The upper surface of the shell (1) is provided with a cover (11), the cover (11) and the shell (1) are detachably connected, the center axis of the cover (11) is provided with an opening (12), the inner edge of the opening (12) is provided with a guide groove (13), and the diameter of the opening (12) is smaller than the side length of the cover (11); the edge of the shell (1) is provided with a plurality of ventilation openings (14).

4. A residential building carbon emissions monitoring system according to claim 3, wherein: The lower surface of the cover (11) is fixedly connected with a nylon brush (15); the nylon brush (15) has bristles with a length in the range of 20-25 mm, the diameter of the bristles of the nylon brush (15) is in the range of 0.3-0.5 mm, the distance between the bristles of the nylon brush (15) is between 0.5-1 mm, the nylon brush (15) is in direct contact with the dustproof mechanism (5); the edge of the shell (1) is provided with a warning structure (16), and the warning structure (16) is electrically connected with the control mechanism.

5. A residential building carbon emissions monitoring system according to claim 4, wherein: The lifting mechanism (2) comprises a spiral blade (21) and a clamping plate (22); one end of the spiral blade (21) is fixedly connected with the clamping plate (22), the clamping plate (22) is a conical structure, and the size of the clamping plate (22) is the same as that of the opening (12); the inside of the spiral blade (21) is a hollow structure.

6. A residential building carbon emissions monitoring system according to claim 5, wherein: The lower end of the spiral blade (21) is a closed loop blade; the maximum static friction force of the outer surface of the spiral blade (21) is less than one tenth of its own weight.

7. A residential building carbon emission monitoring system as claimed in claim 5, wherein: The monitoring mechanism (3) comprises a mobile monitor (31) and a fixed monitor (32); the mobile monitor (31) is clamped in the spiral blade (21), and the fixed monitor (32) is provided in plurality, and the fixed monitor (32) is uniformly distributed on the upper surface of the spiral blade (21) along the spiral blade (21).

8. A residential building carbon emissions monitoring system according to claim 7, wherein: The power mechanism (4) comprises a solar panel (41), an inverter (42), a storage battery (43), a motor (44) and a connecting shaft (45); the solar panel (41) is fixedly connected to the upper surfaces of the cover (11) and the clamping plate (22), the solar panel (41) is electrically connected with the storage battery (43), the storage battery (43) is electrically connected with the inverter (42), and the motor (44) is electrically connected with the storage battery (43) and the control mechanism; the connecting shaft (45) is fixedly connected with the motor (44), and the connecting shaft (45) is fixedly connected with the lower end of the spiral blade (21).

9. A residential building carbon emission monitoring system according to claim 7, wherein: The mobile monitor (31) is electrically connected with the control mechanism, the mobile monitor (31) is provided with a camera function; the fixed monitor (32) is infrared optical sensor, The concentration measurement range is between 0-2000ppm.

10. The residential building carbon emission monitoring system of claim 7, wherein: The dustproof mechanism (5) comprises a frame (51) and a gauze (52); the upper end of the frame (51) is fixedly connected with the clamping plate (22), the gauze (52) is arranged along the frame (51), the upper end of the frame (51) is not provided with the gauze (52), the exposed part of the frame (51) has a length not less than twice the height of the mobile monitor (31), and the lower end of the dustproof mechanism (5) covers the monitoring mechanism (3) in the horizontal direction.

Images