Building group heat supply adjusting method based on infrared temperature measurement of unmanned aerial vehicle

Abstract

The invention discloses a building group heat supply adjusting method based on infrared temperature measurement of an unmanned aerial vehicle. According to the method, temperature distribution of different positions of a building group is inspected in real time through an unmanned aerial vehicle carrying an infrared thermal imager, and temperature data of a building facade and a surrounding environment thereof are obtained by using an infrared temperature measurement technology; and the obtained data is wirelessly transmitted to the data acquisition unit, and the heat supply amount of each single building is automatically adjusted according to the heat demand, so that accurate dynamic temperature control is realized. The system has the advantages that building group temperature monitoring is carried out through automatic inspection of the unmanned aerial vehicle, the problems of inconvenience and errors caused by home-entry monitoring of traditional manual temperature measurement are solved, and the measurement efficiency and precision are improved; an infrared temperature measurement technology is adopted to monitor the external and surrounding environment temperature of the building in real time, and a scientific basis is provided for accurate heat supply adjustment; and through combination with a central intelligent control system, intelligent and refined regulation and control of the building group heat supply system are achieved, energy waste is reduced, and the heat supply efficiency is improved.

Description

Technical Field

[0001] The present invention relates to the fields of drone thermal infrared, building energy conservation and smart heating technology, and in particular to a building complex heating regulation method based on drone infrared temperature measurement, which is suitable for intelligent temperature control management in centralized heating scenarios such as residential communities and commercial buildings. Background Art

[0002] Building temperature measurement is of great significance in the sustainable development of today's society. It helps energy conservation and emission reduction and efficient resource utilization by optimizing energy utilization, ensuring building quality and improving user comfort, and promotes the development of the construction industry in a green, low-carbon and efficient direction, providing strong support for achieving sustainable development goals.

[0003] The current mainstream methods for collecting building temperature include traditional manual measurement, fixed wired measurement, room temperature monitoring based on General Packet Radio Service (GPRS) technology, room temperature monitoring based on wireless Internet of Things (IoT), and infrared remote sensing. The widespread deployment of room temperature collection equipment would incur significant investment costs for heating companies, making it unsuitable for widespread use. Furthermore, battery life issues frequently arise in actual operation of these equipment. When measuring temperature over large areas, ground-based observations can only obtain point-scale surface temperature data, failing to capture detailed information about spatial temperature variations. Infrared remote sensing, on the other hand, utilizes the properties of infrared radiation to measure the temperature of objects. It offers the advantages of high spatial resolution and the absence of direct contact with the object being measured, providing a novel technical support for research related to building room temperature measurement. Therefore, there is an urgent need to develop drone-based room temperature monitoring systems that utilize drones equipped with infrared thermal imaging cameras for non-contact temperature measurement of buildings.

[0004] With the continuous improvement of urbanization in China, the number of heating users and heated areas has increased rapidly. However, the number of vacant homes has increased dramatically, and the number of users experiencing heat supply outages has also increased. This presents new challenges for heating companies. Traditional building heating regulation relies primarily on manual home temperature measurement or fixed sensor networks, which suffer from low efficiency, limited coverage, and data lag. Conducting temperature measurement in building clusters can comprehensively cover all buildings within a region, providing a systematic understanding of the temperature conditions across the entire area. Improving building cluster temperature measurement technology can help identify local or regional issues that may be overlooked when measuring individual buildings. When measuring building temperatures, the lack of comprehensive regional temperature data makes it difficult to accurately control heating or cooling systems, resulting in energy waste or substandard local temperatures. It also makes it difficult to detect temperature anomalies, leading to structural risks due to thermal expansion and contraction or changes in material properties. Uneven indoor temperature distribution in buildings can affect the living and user experience. Furthermore, it is impossible to assess the impact of building clusters on the surrounding urban thermal environment, hindering urban planning and sustainable development.

[0005] Therefore, in order to understand the heating situation of users in a certain area, it is necessary to understand the temperature distribution of the buildings in the area. There is an urgent need to carry out efficient temperature measurement of the buildings.

[0006] Currently, drones can be used to monitor room temperatures by combining them with infrared thermal imaging technology. This technology allows for non-contact temperature measurement of buildings in a target area, effectively measuring surface temperatures. However, while infrared thermal imaging can be used for building temperature measurement, it is often limited to static or localized detection, making it difficult to achieve real-time, dynamic monitoring of the entire building complex.

[0007] Heating control technology can ensure users use heating resources appropriately and legally through regular audits of heating conditions, preventing heat theft and heat leakage. Monitoring heating facilities to check the operating status of heating equipment can effectively identify pipeline leaks and equipment failures. However, because heating system control lacks precise alignment with a building's actual heat demand, energy waste often occurs.

[0008] Therefore, there is an urgent need for an efficient, accurate and non-invasive building temperature monitoring and heating dynamic adjustment method covering different influencing factors. Summary of the Invention

[0009] The technical problem to be solved by the present invention is to provide a building complex heating regulation method based on drone infrared temperature measurement to address the problems in the background technology of limited coverage of traditional temperature measurement and mismatch between heating regulation technology and actual heat demand of buildings.

[0010] The present invention adopts the following technical solutions to solve the above technical problems:

[0011] A building complex heating regulation method based on drone infrared temperature measurement is characterized by achieving refined dynamic management of the heating system through drone inspection, data fusion and intelligent regulation. It includes the following steps:

[0012] Step 1: Use drones to measure the temperature of the target building complex. A swarm of drones equipped with infrared thermal imagers scans the building facade along a preset path, inspecting the temperature distribution at different locations of the building complex in real time, obtaining temperature data on the building facade and its surroundings, and collecting temperature distribution data on windows and wall areas.

[0013] Step 2: Use infrared temperature measurement technology to obtain temperature data of the building facade and its surrounding environment. The obtained data is wirelessly transmitted to the data acquisition unit, and the image pixel temperature is extracted and time series matching is performed. A random forest regression model is constructed by combining drone infrared data, outdoor environmental parameters, and indoor sensor data to predict the indoor temperature of each household. Data normalization and real-time verification: Abnormal data is marked and retaken through edge computing nodes.

[0014] Step 3: Automatically adjust the heating supply of each individual building according to heat demand, thereby achieving precise dynamic temperature control. The central control system automatically adjusts the heating supply of each building unit according to the predicted temperature distribution: the central control system links the heating valves to achieve on-demand distribution.

[0015] Furthermore, as a method for regulating building complex heating based on drone infrared temperature measurement, the present invention selects step 1, using 3 to 5 drones to cover the area where the temperature measurement points are located, with a shooting time of 1–2 hours. The specific shooting time is adjusted appropriately according to the actual building complex facade area and building height. A blackbody-calibrated infrared thermal imager is also used to achieve full coverage, stable, and accurate measurement. A standard blackbody calibrator is used to control the temperature measurement error to ±0.1°C; a handheld thermometer is used to measure the outdoor temperature; a temperature and humidity sensor is placed in the living room for easy access; and an anemometer and solar infrared radiometer are used to record outdoor environmental parameters. Data is recorded once per minute. During drone inspections: Maintain a 5-15m range and a 0° pitch angle based on the actual conditions of the building complex's facade, avoiding obstacles. This temperature measurement condition is the experimentally determined distance and pitch angle for the highest accuracy of drone infrared building complex temperature measurement. Operate in weather conditions without strong winds (<5m / s) and rain or fog, during the 9:00-11:00 or 18:30-19:00 hours to reduce reflection interference.

[0016] The flight route is set from the 1st floor to the top floor, and each floor is photographed in unit order; multiple drones are used to work together and assigned to different buildings or units to ensure full coverage of drone swarm monitoring.

[0017] The heating regulation method for building complexes based on UAV infrared temperature measurement is selected in step 2. A UAV equipped with a thermal imaging camera is used to fly to each household's window along a self-set route. The measuring point distance is 7m, the pitch angle is 0°, and infrared images are taken at the center of the window. The scanning is carried out in layers by building, and 1-2 infrared images are collected for each household, and the environmental parameters are recorded simultaneously.

[0018] Multi-source data fusion is performed. Infrared images captured by thermal imaging cameras are combined with the DJI Thermal SDK and a Python script to extract the temperature of each pixel in the image. This generates a .csv file containing information such as timestamp, building number, unit number, floor, apartment type number, maximum temperature, minimum temperature, and average temperature. Any abnormal data due to occlusion or reflection is flagged by the edge computing node, triggering a retake.

[0019] The data was further processed to perform time series matching, aligning indoor sensor data (such as temperature and humidity) with outdoor data (including infrared temperature, outdoor temperature, wind speed, and solar radiation) by timestamp. Min-Max normalization was performed on building environmental characteristics (including outdoor temperature, wind speed, solar radiation, unit type, and temperature) to generate a unified dataset. A pretrained random forest regression model was loaded and fed with the normalized data to predict the indoor temperature of each household.

[0020] Furthermore, we conduct accuracy assessment and compare the predicted values ​​with the actual values ​​of the sensors for the households where the sensors are installed, and calculate the MSE and R2. The target values ​​are MSE = 0.17 (± 0.005) and R2 = 0.98 (± 0.003), with R2 being the priority. 2 Reach the target value. If deviations occur, check the image quality and environmental parameters, and retake the image or adjust the sensor if necessary. If the accuracy deviation is large (>0.5°C, often seen in unheated side units or top floors), check the image quality for blur or reflections, the environmental parameters for abnormal wind speed or radiation, and the sensor position, and retake or adjust the sensor if necessary.

[0021] Furthermore, image blur and reflection are determined by the characteristic values ​​of pixel parameters extracted from the image. Image quality blur is determined by a decrease in the correlation between the maximum and minimum values ​​of the image pixels, or a decrease in the variance; image reflection is determined by a significant increase in the maximum value or variance of the image pixels. Image parameters vary for different objects captured in different environments, so the magnitude of the change should be determined based on the image parameters of the same object captured in the same environment.

[0022] To further validate the model's adaptability, a random forest model was used to predict indoor temperature. The model was regularly updated to accommodate seasonal variations. Repeated measurements were performed across winter (December–February) and summer (June–August) to verify the model's adaptability to extreme temperatures (-10°C or 35°C). The random forest model was regularly retrained with updated environmental parameters across different seasons to improve its generalization capabilities.

[0023] Furthermore, in the present invention's method for regulating building heating using drone infrared temperature measurement, in step 3, the central control system adjusts the opening of each unit's heating valve based on the temperature distribution. Specifically, the central control system increases the opening of heating valves in low-temperature areas, such as the top floor and side units, while reducing the amount of heat supplied to high-temperature areas to avoid overheating and waste. Combined with heat loss analysis, energy-saving modification recommendations, such as installing additional insulation, are generated. If the output format needs to be temporarily adjusted, it can be dynamically generated using Python pandas.

[0024] Real-time monitoring uses a cloud platform to stream drone and sensor data, updated every five minutes. A web dashboard built using Python Dash or Flask generates dynamic temperature distribution. Thresholds are set to automatically send alerts to property management when the indoor temperature drops below 18°C ​​or exceeds 25°C. Heat distribution is optimized to reduce heat loss in top-floor units or adjust heating valves. Residents with high heat loss are identified and personalized recommendations are provided, including installing thermal film or replacing triple-glazed windows.

[0025] Furthermore, the collected data is stored on a local server and integrated into the building management system. A summary report is generated for each building, including a temperature distribution map, a list of abnormal residents, energy-saving recommendations, and data consistency verification, ensuring that the deviation between the predicted value and the sensor value is less than 0.5°C. The temperature data is compatible with the community heating system.

[0026] Furthermore, during the flight of the drone, the infrared thermal imaging camera it carries is combined with the geographic information system (GIS) and big data analysis. If there is a fault point, infrared thermal imaging technology can be used to quickly locate the pipeline leakage point or abnormal hotspot, quickly feed back the information to the control center and transmit the on-site image in real time, thereby improving the emergency response speed.

[0027] Compared with the prior art, the present invention adopts the above technical solution and has the following technical advantages:

[0028] 1. This invention achieves efficient and accurate non-contact temperature measurement using drones, can cover the entire area, avoid interference from entering homes, improve measurement efficiency by more than 80%, and achieve a temperature prediction accuracy of ±0.5°C;

[0029] 2. This invention dynamically adjusts the heating supply of buildings based on real-time data, achieving intelligent and precise control based on user characteristics, which can reduce energy consumption by 10%-15%;

[0030] 3. The present invention can quickly identify and feedback fault points through infrared thermal imaging technology, thereby improving the emergency response speed during the heating season. BRIEF DESCRIPTION OF THE DRAWINGS

[0031] Figure 1 The figure is a flowchart of the heating regulation method based on infrared temperature measurement of a drone according to the present invention. DETAILED DESCRIPTION

[0032] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of them. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts shall fall within the scope of protection of the present invention.

[0033] Reference Figure 1A building complex heating regulation method based on drone infrared temperature measurement is characterized by comprising the following steps:

[0034] Step 1: Use drones to measure the temperature of the target building complex. A swarm of drones equipped with infrared thermal imagers scans the building facade along a preset path, inspecting the temperature distribution at different locations of the building complex in real time, obtaining temperature data on the building facade and its surroundings, and collecting temperature distribution data on windows and wall areas.

[0035] Step 2: Use infrared temperature measurement technology to obtain temperature data of the building facade and its surrounding environment. The obtained data is wirelessly transmitted to the data acquisition unit, and the image pixel temperature is extracted and time series matching is performed. A random forest regression model is constructed by combining drone infrared data, outdoor environmental parameters, and indoor sensor data to predict the indoor temperature of each household. Data normalization and real-time verification: Abnormal data is marked and retaken through edge computing nodes.

[0036] Step 3: Automatically adjust the heating supply of each individual building according to heat demand, thereby achieving precise dynamic temperature control. The central control system automatically adjusts the heating supply of each building unit according to the predicted temperature distribution: the central control system links the heating valves to achieve on-demand distribution.

[0037] In a further embodiment, step 1 further includes placing the temperature and humidity sensor in a conveniently accessible living room. The test point is required to be 1.4 meters high, 1 meter from a wall, and away from radiators, air conditioners, or direct sunlight. An anemometer and a solar infrared radiometer are used to record outdoor environmental parameters. After calibration, the instruments are placed in a shaded, ventilated area to avoid direct sunlight or in a portable case to avoid temperature fluctuations.

[0038] In a further embodiment, step 2 also includes: setting the flight path to automatically ascend from the 1st to the 14th floor when surveying high-rise buildings, and from the 1st to the 8th floor when surveying low-rise buildings, capturing each floor in unit order; using multiple drones to coordinate operations, assigning them to different buildings or units to ensure full coverage of the drone swarm. DJI Pilot monitors the location in real time to avoid collisions, and maintains a safe distance of 50 meters between each drone.

[0039] In a further embodiment, the infrared images captured by the thermal imaging camera are used to extract the temperature of each pixel in the image through the DJI Thermal SDK combined with a Python script to generate a .csv file. The data information includes: timestamp, building number, unit number, floor, apartment number, maximum temperature, minimum temperature, and average temperature.

[0040] In a further embodiment, the data is processed to complete time series matching, aligning indoor sensor data such as temperature and humidity with outdoor data: infrared temperature, outdoor temperature, wind speed, and solar radiation by timestamp; Min-Max normalization is performed on building environmental features: outdoor temperature, wind speed, solar radiation, apartment number, and temperature features to unify the data set. Load the pre-trained random forest regression model, MSE = 0.1723, R 2 =0.9777, stored as a .joblib file, input the normalized data, and predict the indoor temperature of each household;

[0041] In a further embodiment, the accuracy is determined by comparing the predicted values ​​with the actual values ​​measured by the sensors for the households where the sensors are installed, and calculating the MSE and R 2 The target value is MSE = 0.17 (± 0.005), R 2 =0.98(±0.003), giving priority to ensuring R 2 Reach the target value. When deviation occurs, check the image quality and environmental parameters, and retake or adjust the sensor if necessary. If the accuracy deviation is large (>0.5℃, commonly seen in unheated side doors or top floors), check whether the image quality is blurred or reflective, whether the environmental parameters have abnormal wind speed (>5m / s), and the sensor position. If necessary, retake or adjust the sensor to keep it away from the heat source. Image quality is determined by the characteristic values ​​of the pixel parameters extracted from the image. For image quality blur, it is determined by the lower correlation between the maximum and minimum values ​​of the image pixels, or the smaller variance; for image reflection, it is determined by the significant increase in the maximum value of the image pixels and the significant increase in the variance. The image parameters of different objects shot in different environments are different, so the determination of the size of the change value needs to be based on the image parameters of the same environment and the same object.

[0042] In a further embodiment, when verifying the adaptability of the model, a random forest model is used to predict the indoor temperature, and the model is regularly updated to adapt to seasonal changes. The cross-seasonal application is repeated in winter (December-February) and summer (June-August) to verify the adaptability of the model in extreme temperatures (-10°C or 35°C). The random forest model is regularly retrained by updating the environmental parameters of different seasons to improve the generalization ability of the model, with the goal of R 2 >0.98.

[0043] In a further embodiment, the central control system adjusts the opening of each unit's heating valve based on temperature distribution. Specifically, it increases the opening of heating valves in low-temperature areas, typically located on the top floor and in side units, while reducing the amount of heat supplied to high-temperature areas to avoid overheating and waste. Combined with heat loss analysis, often focusing on issues like window leaks, it generates energy-saving retrofit recommendations, including the installation of additional insulation. If the output format needs to be temporarily adjusted, such as when the property management requires specific fields, it can be dynamically generated using Python pandas.

[0044] In a further embodiment, real-time monitoring is achieved through a cloud platform that streams drone and sensor data, updating it every five minutes. A web dashboard built using Python Dash or Flask generates a dynamic temperature profile. Thresholds are set to automatically send alerts to property management via email or API push when the indoor temperature drops below 18°C ​​or exceeds 25°C. Heat distribution is optimized to reduce heat loss in top-floor units or adjust heating valves. Residents with high heat loss are identified and personalized recommendations are provided, including installing thermal film or replacing triple-glazed windows.

[0045] In a further embodiment, the collected data is saved to a local server and integrated into the building management system to generate a summary report for each building, including a temperature distribution heat map, a list of abnormal residents (when the temperature is <18°C or >25°C, the reasons include insufficient heating and window leakage), and energy-saving suggestions; installation of insulation layer, upgrading of double-layer Low-E glass, and proofreading of data consistency, that is, the deviation between the predicted value and the sensor value is <0.5°C.

[0046] During the flight of the drone, the infrared thermal imaging camera it carries is combined with the geographic information system (GIS) and big data analysis. If there is a fault point, infrared thermal imaging technology is used to quickly locate the pipeline leakage point or abnormal hotspot, quickly feed back the information to the control center and transmit the on-site image in real time, thereby improving the emergency response speed.

[0047] Finally, a few points should be explained: First, in the description of this application, it should be noted that, unless otherwise specified or limited, the terms "mounted," "connected," and "connected" should be understood in a broad sense, and may refer to mechanical or electrical connections, internal communication between two components, or direct connection. "Up," "down," "left," and "right" are only used to indicate relative positional relationships. When the absolute positions of the objects being described change, the relative positional relationships may also change.

[0048] Secondly: The drawings of the embodiments disclosed in the present invention only involve structures related to the embodiments disclosed in the present invention. Other structures may refer to conventional designs. The same embodiment and different embodiments of the present invention may be combined with each other without conflict.

[0049] Finally: The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A building complex heating regulation method based on drone infrared temperature measurement, characterized in that: The following steps are involved: Step 1: Use drones to measure the temperature of the target building complex. A swarm of drones equipped with infrared thermal imagers scans the building facade along a preset path, inspecting the temperature distribution at different locations of the building complex in real time, obtaining temperature data on the building facade and its surroundings, and collecting temperature distribution data on windows and wall areas. Step 2: Use infrared temperature measurement technology to obtain temperature data of the building facade and its surrounding environment. The obtained data is wirelessly transmitted to the data acquisition unit, and the image pixel temperature is extracted and time series matching is performed. The indoor temperature of each household is predicted by combining the drone infrared data, outdoor environmental parameters, and indoor sensor data. Data normalization and real-time verification are performed: abnormal data is marked and a retake mechanism is triggered through the edge computing node. Step 3: Automatically adjust the heating supply of each individual building according to the heat demand, thereby realizing dynamic temperature control. The central control system automatically adjusts the heating supply of each building unit according to the predicted temperature distribution: the central control system links the heating valves to achieve on-demand distribution.

2. The method for regulating heat supply of a building complex based on infrared temperature measurement by drone according to claim 1 is characterized in that: In step 1, 3 to 5 drones are used to cover the area where the temperature measurement points are located. The shooting time lasts for 1-2 hours and the drones are equipped with infrared thermal imagers. During the inspection, the drones maintain a range of 5-15m and a pitch angle of 0° to avoid obstacles. Work in weather conditions with wind speed less than 5m / s and no rain or fog, between 9:00-11:00 or 18:30-19:00 to reduce reflection interference.

3. The method for regulating heat supply of a building complex based on infrared temperature measurement by drone according to claim 1, characterized in that: In step 2, multi-source data fusion is performed. Infrared images captured by the thermal imaging camera are combined with the DJI Thermal SDK and a Python script to extract the temperature of each pixel in the image. The data information includes: timestamp, building number, unit number, floor, apartment number, maximum temperature, minimum temperature, and average temperature; outdoor environmental parameters include infrared temperature, outdoor temperature, wind speed, and solar radiation.

4. The method for regulating heat supply of buildings based on infrared temperature measurement by drone according to claim 1, characterized in that: In step 2, a random forest model is used to predict indoor temperature. The model is regularly updated to adapt to seasonal changes; the random forest model is retrained every three months.

5. The method for regulating heat supply of buildings based on infrared temperature measurement by drone according to claim 1, characterized in that: In step 2, the data is processed to complete time series matching and align the indoor sensor data with the outdoor data by timestamp; the building environmental characteristics are subjected to Min-Max normalization to generate a unified data set; and a pre-trained random forest regression model is loaded, the normalized data is input, and the indoor temperature of each household is predicted.

6. The method for regulating heat supply of buildings based on infrared temperature measurement by drone according to claim 1, characterized in that: In step 3, the central control system adjusts the opening of the heating valve of each unit according to the temperature distribution, specifically: increasing the opening of the heating valve in the low-temperature area; reducing the heat supply to the high-temperature area to avoid overheating waste; or generating energy-saving transformation suggestions based on heat loss analysis.

7. The method for regulating heat supply of buildings based on infrared temperature measurement by drone according to claim 1, characterized in that: In step 3, real-time monitoring is performed by uploading drone and sensor data streams through a cloud platform to generate a dynamic temperature distribution. When the indoor temperature is less than 18°C ​​or greater than 25°C, an early warning is automatically sent to the property management to optimize heat distribution.

8. The method for regulating heat supply of buildings based on infrared temperature measurement by drone according to claim 1, characterized in that: In step 1, the flight route is set from the first floor to the top floor, and each floor is photographed in unit order; multiple drones are used to work together and assigned to different buildings or units to ensure full coverage of drone group monitoring; By flying to each window along a self-set route, the measuring point distance is maintained at 5-15m according to the actual building facade conditions, the pitch angle is 0°, and infrared images are taken at the center of the window. Scanning is performed in layers by building, and 1-2 infrared images are collected for each household, while environmental parameters are recorded simultaneously.

9. The method for regulating heat supply of buildings based on infrared temperature measurement by drone according to claim 9, characterized in that: In step 2, the accuracy is determined by comparing the predicted value with the actual value of the sensor for the household where the sensor is installed, and calculating the MSE or R 2 ; The target value is MSE less than 0.17 (±0.005), R 2 >0.98(±0.003), priority is given to ensuring R 2 Reach the target value; when deviation occurs, check the image quality and environmental parameters. When the accuracy deviation is greater than 0.5℃, check whether the image quality is blurred or reflective, whether there are abnormal environmental parameters such as wind speed greater than 5m / s, and whether there is deviation from the sensor position. If necessary, retake the picture or adjust the sensor.

10. The method for regulating heat supply of buildings based on infrared temperature measurement by drone according to claim 1, characterized in that: In step 2, the characteristic values ​​of the pixel parameters extracted from the image specifically refer to the maximum value, minimum value and variance.

Images