Urban residential quarter energy consumption management system based on Internet of Things

Abstract

The invention relates to the technical field of energy consumption management, in particular to an urban residential quarter energy consumption management system based on the Internet of Things, which comprises an image acquisition unit, an energy consumption identification unit, an energy consumption statistical unit, a cost comparison unit, an anomaly detection unit, a data display unit and an energy saving prediction unit, specifically, the image acquisition unit is used for performing image acquisition on an electricity meter and a water meter in a target residential quarter according to a preset sampling period based on a camera so as to obtain an electricity meter image and a water meter image of each user, and transmitting the electricity meter image and the water meter image to the energy consumption identification unit. According to the system, images of the electricity meter and the water meter are collected regularly through the camera, accurate recognition is conducted through the energy consumption recognition unit, errors caused by manual meter reading can be avoided through the system, the accuracy of data collection is ensured, it is ensured that electricity consumption and water consumption data of a user can be updated in time through real-time monitoring, and therefore delay or errors in manual recording are avoided.

Description

Technical Field

[0001] The present invention relates to the technical field of energy consumption management, and in particular to an energy consumption management system for urban residential areas based on the Internet of Things. Background Art

[0002] In the architecture of the Internet of Things, cameras are usually used as one of the "perception layer" devices to collect image and video data and transmit them to the Internet of Things platform for analysis and processing. Cameras are connected to other devices through the network and can transmit monitoring data in real time to form a complete intelligent system.

[0003] Traditional systems usually rely on manual meter reading, which is prone to data entry errors or omissions, and there may be a time lag in the meter reading process, which cannot reflect the user's electricity or water consumption in real time, which not only leads to inaccurate data, but also affects the subsequent cost accounting and resource allocation, which is easy to cause inconvenience to users and managers; and in traditional systems, fee settlement usually relies on manual verification, calculation and adjustment, which is easy to cause errors due to manual operation. Errors may occur in user fee settlement, such as missed or repeated billing, and the settlement cycle may be long, and it is impossible to achieve real-time or timely fee updates, which brings difficulties to management; and traditional systems usually lack real-time energy consumption monitoring functions, and cannot obtain electricity and water consumption data of each user in real time. Data is usually collected periodically and cannot reflect the energy consumption at a certain moment in real time, resulting in the inability to detect abnormalities in time or conduct effective management; and most traditional systems lack in-depth analysis and prediction functions of historical data, and cannot predict the energy-saving potential of users, nor can they provide personalized energy-saving suggestions. Users' energy consumption behavior is difficult to be fully understood and analyzed, and managers cannot take targeted energy-saving measures or provide targeted energy-saving guidance to users. Summary of the invention

[0004] The technical problem to be solved by the present invention is to overcome the shortcomings of the above-mentioned prior art and provide an urban residential area energy consumption management system based on the Internet of Things.

[0005] The technical solution adopted to solve the above technical problems is: an energy consumption management system for urban residential areas based on the Internet of Things, including an image acquisition unit, an energy consumption recognition unit, an energy consumption statistics unit, a cost comparison unit, an anomaly detection unit, a data display unit and an energy saving prediction unit, specifically: The image acquisition unit is used to acquire images of electric meters and water meters in the target residential area based on the camera according to a preset sampling period to obtain electric meter images and water meter images of each user, and transmit the electric meter images and water meter images to the energy consumption identification unit; The energy consumption identification unit is used to receive the electric meter image and the water meter image transmitted by the image acquisition unit, and perform image recognition on the electric meter image and the water meter image based on a pre-trained image recognition model to obtain the cumulative electricity consumption corresponding to the electric meter image and the cumulative water consumption corresponding to the water meter image, and obtain the current electricity consumption and the current water consumption based on the cumulative electricity consumption and the cumulative water consumption in the current acquisition cycle and the cumulative electricity consumption and the cumulative water consumption in the previous acquisition cycle, that is, obtain the electricity consumption and the water consumption of each user in the target residential community, and transmit the electricity consumption and the water consumption of each user to the energy consumption statistics unit and the cost comparison unit.

[0006] Preferably, the fee comparison unit is used to receive the electricity consumption and water consumption of each user transmitted by the energy consumption identification unit, and obtain the electricity cost and water cost of each user based on the preset electricity unit price and water unit price, obtain the prepaid electricity cost and prepaid water cost of each user based on the electricity cost and water cost of each user, compare the prepaid electricity cost and prepaid water cost of each user with the pre-acquired electricity supply bill and water supply bill to obtain the difference in electricity cost and water cost of each user, and make supplements and refunds to each user based on the difference in electricity cost and water cost of each user.

[0007] Preferably, the energy consumption statistics unit is used to receive the electricity consumption and water consumption of each user transmitted by the energy consumption identification unit, and sort the electricity consumption and water consumption of each user according to a preset collection period to obtain the electricity consumption sequence and water consumption sequence of each user, obtain the electricity consumption sequence and water consumption sequence of each area in the target residential community based on the electricity consumption sequence and water consumption sequence of each user, transmit the electricity consumption sequence and water consumption sequence of each user to the abnormality detection unit, and transmit the electricity consumption sequence and water consumption sequence of each area in the target residential community to the data display unit.

[0008] Preferably, the abnormality detection unit is used to receive the electricity consumption sequence and water consumption sequence of each user transmitted by the energy consumption statistics unit, and cluster the electricity consumption sequence and water consumption sequence of each user to obtain various electricity consumption conditions and various water consumption conditions, and cluster the electricity consumption sequence and water consumption sequence of each user again to determine the electricity consumption benchmarks of various electricity consumption conditions and the water consumption benchmarks of various water consumption conditions, and judge whether the electricity consumption sequence and water consumption sequence of each user have abnormal electricity consumption and water use based on the electricity consumption benchmarks of various electricity consumption conditions and the water consumption benchmarks of various water consumption conditions; if the electricity consumption and water use of each user are abnormal, the various electricity consumption conditions and various water consumption conditions as well as the electricity consumption benchmarks of various electricity consumption conditions and the water consumption benchmarks of various water conditions are transmitted to the energy-saving prediction unit.

[0009] Preferably, the energy-saving prediction unit is used to receive various electricity usage conditions and various water usage conditions as well as electricity usage benchmarks of various electricity usage conditions and water usage benchmarks of various water usage conditions transmitted by the abnormality detection unit, and predict the energy-saving capabilities of users with abnormal electricity and water usage based on the various electricity usage conditions and various water usage conditions as well as the electricity usage benchmarks of various electricity conditions and water usage benchmarks of various water conditions, and transmit the energy-saving capabilities of users with abnormal electricity and water usage to the data display unit.

[0010] Preferably, the data display unit is used to receive the electricity consumption sequence and water consumption sequence of each area in the target residential community transmitted by the energy consumption statistics unit and the energy-saving capability of users with abnormal electricity and water consumption transmitted by the energy-saving prediction unit, obtain energy-saving electricity consumption and energy-saving water consumption based on the energy-saving capability of users with abnormal electricity and water consumption, display the electricity consumption sequence and water consumption sequence of each area in the target residential community, and display the energy-saving electricity consumption and energy-saving water consumption of users with abnormal electricity and water consumption.

[0011] Preferably, the image recognition model includes a regional positioning module, a background filtering module, a correction module and a digital recognition module, wherein the regional positioning module is used to perform regional positioning on the visible area of ​​the electric meter image and the water meter image to obtain the visible area image, the background filtering module is used to detect the visible area image to filter the background information of the visible area image, thereby obtaining the digital image of the visible area image, the correction module is used to correct the digital image to obtain a digital image at a standard angle, and the digital recognition module is used to perform digital recognition on the digital image at a standard angle to obtain the number corresponding to the digital image at the standard angle, thereby obtaining the cumulative electricity consumption corresponding to the electric meter image and the cumulative water consumption corresponding to the water meter image.

[0012] Preferably, the region positioning module adopts a YOLOv4-tiny model, the background filtering module adopts an improved DMPNet model, and the digital recognition module adopts a convolutional neural network.

[0013] Preferably, the background filtering module is based on inserting two control points along the upper and lower boundaries of the display area image to improve the loss function of the DMPNet model, wherein the improved loss function is as follows: ; in, represents the position loss of 8 fitting points, and Represents the predicted value and true value of a certain point, represents the sum of the centrality scores of the 8 fitting points, represents the error between the predicted value and the true value at a certain point, and .

[0014] Preferably, the digital image is corrected to obtain a digital image at a standard angle, comprising: Acquire a corner point set and a control point set of the digital image, and calculate a minimum convex hull of the corner point set and the control point set, wherein the minimum convex hull represents a minimum polygon enclosed by the corner point set and the control point set; Calculate the minimum circumscribed rectangle of the minimum convex hull, and obtain the width and height of the minimum circumscribed rectangle; Traversing each point in the digital image, obtaining a first distance between each point and the left border of the digital image, and calculating a ratio of the first distance to a width of the minimum circumscribed rectangle; The upper and lower boundary points of the digital image are determined based on the ratio of the first distance to the minimum circumscribed rectangle width, the corner point set and the control point set, wherein the expressions of the upper and lower boundaries are as follows: ; in, Indicates the upper and lower boundaries, and Represents the left and right corners, and represents the control point, Indicates the ratio of the first distance to the width of the minimum enclosing rectangle; Traversing each point in the digital image, obtaining a second distance between each point and the upper boundary of the digital image, and calculating a ratio of the second distance to the height of the minimum circumscribed rectangle; The corresponding position of each point in the digital image is calculated based on the ratio of the upper and lower boundary points of the digital image to the second distance and the height of the minimum circumscribed rectangle, thereby obtaining a digital image at a standard angle, wherein the expression of the corresponding position in the digital image is as follows: ; in, represents the corresponding position of each point in the digital image, and Represents the upper and lower boundary points of the digital image, Indicates the ratio of the second distance to the height of the minimum enclosing rectangle.

[0015] Preferably, clustering the electricity consumption sequence and water consumption sequence of each user includes: The number of clusters of the electricity consumption sequence and the water consumption sequence is set, and a sample point is randomly selected as the initial cluster center; Calculate the shortest distance from each sample point to the existing cluster center, calculate the probability of each sample point becoming the next cluster center, and take the sample point with the maximum probability as the next cluster center; Repeat the above operation until a preset number of initial cluster centers are selected; Calculate the Euclidean distance between the sample point and the cluster center, compare the Euclidean distance between each sample point and each cluster center, and divide each sample point into the class where the nearest cluster center is located; Calculate the sample mean points of each category, update the cluster center, and repeat the above operations until the position of each cluster center no longer changes.

[0016] Preferably, the calculation formula for the energy-saving capability of the user with abnormal electricity and water consumption is as follows: ; in, Indicates the energy-saving capabilities of users with abnormal electricity and water usage, Indicates Classification of electricity usage conditions and various water usage conditions Sampling period The electricity and water consumption at each moment, Indicates Types of electricity and water usage The electricity and water usage standards at the moment, Indicates the total types of electrical conditions and various water-using conditions. Indicates the total number of moments in the sampling period, Indicates the total number of sampling periods.

[0017] The beneficial effects of the present invention are as follows: (1) The present invention uses a camera to regularly collect images of the electricity meter and the water meter, and uses an energy consumption identification unit to accurately identify them. The system can avoid errors in manual meter reading and ensure the accuracy of data collection. Real-time monitoring ensures that the user's electricity and water consumption data can be updated in a timely manner, thereby avoiding delays or errors in manual records. The real-time collected electricity and water consumption data are automatically compared with the preset unit price, and the user's electricity and water charges are automatically calculated. The difference between the prepaid fee and the actual fee is used to make up or refund the fee, which greatly simplifies the cost accounting process and improves work efficiency. This avoids manual operation errors and ensures the accuracy of fee settlement; (2) The present invention can count and analyze the electricity and water consumption of each user, helping managers understand the energy consumption of different areas. This provides data support for energy efficiency planning, allowing managers to reasonably allocate and manage resources according to energy consumption trends in different areas, avoiding energy waste and improving overall energy efficiency. In addition, the anomaly detection unit can identify abnormal electricity or water consumption behaviors through cluster analysis of energy consumption data. For example, if a user's electricity consumption suddenly increases, the system will automatically issue an alarm to prompt managers and users to check and take measures. This can timely detect water leaks, equipment failures or other abnormal energy usage phenomena, and prevent energy waste or equipment damage; (3) The present invention can predict which users have energy-saving potential based on historical electricity and water usage data and anomaly detection. By analyzing the energy usage behaviors of different users, the system can identify which users may waste energy and provide energy-saving suggestions. This not only helps users reduce unnecessary energy waste, but also improves the energy efficiency of the entire community. The data display unit can intuitively display the electricity and water usage of each area and user, and provide energy-saving prediction information. Users can view their energy usage status in real time, discover potential energy-saving opportunities, and enhance energy-saving awareness. Through user participation in energy-saving management, energy consumption can be more effectively reduced. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] Figure 1 A schematic diagram of the system architecture of an overall system in an embodiment of the present invention.

[0019] Figure numerals: 1. Image acquisition unit; 2. Energy consumption identification unit; 3. Energy consumption statistics unit; 4. Cost comparison unit; 5. Abnormality detection unit; 6. Data display unit; 7. Energy saving prediction unit. DETAILED DESCRIPTION

[0020] Embodiment 1, as Figure 1 As shown, the present invention proposes an urban residential area energy consumption management system based on the Internet of Things, including an image acquisition unit 1, an energy consumption recognition unit 2, an energy consumption statistics unit 3, a cost comparison unit 4, an abnormality detection unit 5, a data display unit 6 and an energy saving prediction unit 7, specifically: The image acquisition unit 1 is used to acquire images of electric meters and water meters in the target residential area based on a camera according to a preset sampling period to obtain electric meter images and water meter images of each user, and transmit the electric meter images and water meter images to the energy consumption identification unit 2; The energy consumption identification unit 2 is used to receive the electric meter image and the water meter image transmitted by the image acquisition unit 1, and perform image recognition on the electric meter image and the water meter image based on a pre-trained image recognition model to obtain the cumulative electricity consumption corresponding to the electric meter image and the cumulative water consumption corresponding to the water meter image, and obtain the current electricity consumption and the current water consumption based on the cumulative electricity consumption and the cumulative water consumption of the current acquisition cycle and the cumulative electricity consumption and the cumulative water consumption of the previous acquisition cycle, that is, obtain the electricity consumption and the water consumption of each user in the target residential community, and transmit the electricity consumption and the water consumption of each user to the energy consumption statistics unit 3 and the cost comparison unit 4.

[0021] In the present invention, camera acquisition refers to the use of a camera device to capture images of a target (electricity meter, water meter). The camera can be an ordinary digital camera or a dedicated monitoring device, usually installed in a place where the electricity meter and water meter can be clearly captured; the sampling period refers to the frequency or interval of data acquisition within a certain time interval. In this scenario, the camera will capture images of the electricity meter and water meter according to a predetermined time interval (such as daily, weekly or monthly). This period can be flexibly set according to actual needs; the cumulative electricity consumption and cumulative water consumption refer to the total electricity consumption and water consumption recorded by the electricity meter and water meter from the installation of the equipment to the current moment.

[0022] In an optional embodiment, the cost comparison unit 4 is used to receive the electricity consumption and water consumption of each user transmitted by the energy consumption identification unit 2, and obtain the electricity cost and water cost of each user based on the preset electricity unit price and water unit price, obtain the prepaid electricity cost and prepaid water cost of each user based on the electricity cost and water cost of each user, compare the prepaid electricity cost and prepaid water cost of each user with the pre-acquired electricity supply bill and water supply bill to obtain the difference in electricity cost and water cost of each user, and make supplements and refunds to each user based on the difference in electricity cost and water cost of each user.

[0023] In an optional embodiment, the energy consumption statistics unit 3 is used to receive the electricity consumption and water consumption of each user transmitted by the energy consumption identification unit 2, and sort the electricity consumption and water consumption of each user according to a preset collection period to obtain the electricity consumption sequence and water consumption sequence of each user, obtain the electricity consumption sequence and water consumption sequence of each area in the target residential community based on the electricity consumption sequence and water consumption sequence of each user, transmit the electricity consumption sequence and water consumption sequence of each user to the abnormality detection unit 5, and transmit the electricity consumption sequence and water consumption sequence of each area in the target residential community to the data display unit 6.

[0024] In an optional embodiment, the abnormality detection unit 5 is used to receive the electricity consumption sequence and water consumption sequence of each user transmitted by the energy consumption statistics unit 3, and cluster the electricity consumption sequence and water consumption sequence of each user to obtain various electricity consumption conditions and various water consumption conditions, and cluster the electricity consumption sequence and water consumption sequence of each user again to determine the electricity consumption benchmarks of various electricity consumption conditions and the water consumption benchmarks of various water consumption conditions, and judge whether the electricity consumption sequence and water consumption sequence of each user are abnormal in electricity and water consumption based on the electricity consumption benchmarks of various electricity consumption conditions and the water consumption benchmarks of various water consumption conditions. If the electricity consumption and water consumption of each user are abnormal, the various electricity consumption conditions and various water consumption conditions as well as the electricity consumption benchmarks of various electricity consumption conditions and the water consumption benchmarks of various water conditions are transmitted to the energy-saving prediction unit 7.

[0025] It should be noted that electricity usage conditions refer to the state or pattern of electricity usage of users under different conditions. Different users may have different electricity usage behaviors. For example, electricity consumption is higher during peak hours. Some users may use electricity all day, while others use electricity during specific periods. Through clustering, users can be divided into several typical electricity usage conditions, and each condition reflects a typical electricity usage pattern. Water use conditions are similar to electricity usage conditions, which refer to the state or pattern of water use of users under different conditions. For example, some users use more water in the morning and evening, while others may use more water at noon and at night. Through cluster analysis, different types of water use conditions can be identified. Abnormal electricity consumption refers to the user's electricity consumption significantly deviating from his or her electricity consumption benchmark or normal electricity consumption range. Abnormal electricity consumption may be caused by various reasons (such as electrical failure, electricity theft, equipment damage, etc.).

[0026] In an optional embodiment, the energy-saving prediction unit 7 is used to receive various electricity usage conditions and various water usage conditions as well as electricity usage benchmarks of various electricity usage conditions and water usage benchmarks of various water usage conditions transmitted by the abnormality detection unit 5, and predict the energy-saving capabilities of users with abnormal electricity and water usage based on the various electricity usage conditions and various water usage conditions as well as the electricity usage benchmarks of various electricity usage conditions and water usage benchmarks of various water conditions, and transmit the energy-saving capabilities of users with abnormal electricity and water usage to the data display unit 6.

[0027] In an optional embodiment, the data display unit 6 is used to receive the electricity consumption sequence and water consumption sequence of each area in the target residential community transmitted by the energy consumption statistics unit 3 and the energy-saving capabilities of users with abnormal electricity and water usage transmitted by the energy-saving prediction unit 7, obtain energy-saving electricity consumption and energy-saving water consumption based on the energy-saving capabilities of users with abnormal electricity and water usage, display the electricity consumption sequence and water consumption sequence of each area in the target residential community, and display the energy-saving electricity consumption and energy-saving water consumption of users with abnormal electricity and water usage.

[0028] Embodiment 2: An urban residential area energy consumption management system based on the Internet of Things proposed by the present invention, compared with embodiment 1, this embodiment also includes: an image recognition model includes a regional positioning module, a background filtering module, a correction module and a digital recognition module, wherein the regional positioning module is used to perform regional positioning on the visible area of ​​the electric meter image and the water meter image to obtain the visible area image, the background filtering module is used to detect the visible area image to filter the background information of the visible area image, thereby obtaining a digital image of the visible area image, the correction module is used to correct the digital image to obtain a digital image at a standard angle, and the digital recognition module is used to perform digital recognition on the digital image at a standard angle to obtain the number corresponding to the digital image at the standard angle, thereby obtaining the cumulative electricity consumption corresponding to the electric meter image and the cumulative water consumption corresponding to the water meter image.

[0029] In an optional embodiment, the region positioning module adopts the YOLOv4-tiny model, the background filtering module adopts the improved DMPNet model, and the digital recognition module adopts the convolutional neural network.

[0030] It should be noted that the YOLOv4-tiny model is a target detection algorithm that is widely used for object recognition and positioning in images. The core feature of the YOLO series of models is to transform the target detection task into a regression problem, and use a neural network model to simultaneously predict the location and category of the object in a single forward propagation; YOLOv4-tiny is a simplified version of YOLOv4. Compared with YOLOv4, the YOLOv4-tiny model is lighter, faster, and consumes less computing resources. It is suitable for devices with weaker hardware performance or application scenarios with high real-time requirements. Due to its small size, it is usually used in embedded devices or situations that require efficient reasoning; DMPNet (Deep Multi-scale Perception A convolutional neural network (CNN) is a deep learning model that is commonly used for background separation and processing of images. It can process images on feature maps of different scales and adapt to targets or scenes of different sizes. An improved DMPNet usually means that the original DMPNet model has been optimized or enhanced for specific tasks or data sets. The improvements may include modifying the network architecture, adding new layers, improving the loss function, enhancing data preprocessing, or performing more complex feature fusion to improve the performance of the model in specific tasks (such as background filtering). A convolutional neural network (CNN) is a deep learning model that is particularly suitable for image processing. CNN simulates the biological visual system and uses convolution operations to extract local features in the image. Then, it gradually transforms the local features into high-level features through operations such as pooling and full connection, which are finally used for classification, recognition, or regression tasks.

[0031] In an optional embodiment, the background filtering module inserts two control points along the upper and lower boundaries of the display area image to improve the loss function of the DMPNet model, wherein the improved loss function is as follows: ; in, represents the position loss of 8 fitting points, and Represents the predicted value and true value of a certain point, represents the sum of the centrality scores of the 8 fitting points, represents the error between the predicted value and the true value at a certain point, and .

[0032] It should be noted that control point is a commonly used term in image processing and computer vision, which refers to specific positions or points selected in the image. These points are used to control the behavior of the image processing algorithm or optimize the target. By selecting control points, the algorithm can perform fine processing or adjustments on certain specific areas of the image.

[0033] In an optional embodiment, correcting the digital image to obtain a digital image at a standard angle includes: Obtaining a corner point set and a control point set of a digital image, and calculating the minimum convex hull of the corner point set and the control point set, wherein the minimum convex hull represents the minimum polygon enclosed by the corner point set and the control point set; Calculate the minimum enclosing rectangle of the minimum convex hull, and obtain the width and height of the minimum enclosing rectangle; Traversing each point in the digital image, obtaining a first distance between each point and the left border of the digital image, and calculating a ratio between the first distance and a minimum circumscribed rectangle width; The upper and lower boundary points of the digital image are determined based on the ratio of the first distance and the minimum circumscribed rectangle width, the corner point set and the control point set, wherein the expressions of the upper and lower boundaries are as follows: ; in, Indicates the upper and lower boundaries, and Represents the left and right corners, and represents the control point, Indicates the ratio of the first distance to the width of the minimum enclosing rectangle; Traversing each point in the digital image, obtaining a second distance between each point and the upper boundary of the digital image, and calculating a ratio between the second distance and a height of a minimum circumscribed rectangle; The corresponding position of each point in the digital image is calculated based on the ratio of the upper and lower boundary points of the digital image to the second distance and the height of the minimum circumscribed rectangle, thereby obtaining a digital image at a standard angle, wherein the expression of the corresponding position in the digital image is as follows: ; in, represents the corresponding position of each point in the digital image, and Represents the upper and lower boundary points of the digital image, Indicates the ratio of the second distance to the height of the minimum enclosing rectangle.

[0034] It should be noted that the minimum enclosing rectangle is the smallest rectangle that encloses a point set or convex hull, and its sides are parallel to or diagonally aligned with the coordinate axes. For digital images, the minimum enclosing rectangle can be used to represent the area range of the image or determine the area that needs to be corrected during the standardization process; the corresponding position in the image means that during the image correction process, each point in the image is mapped to a new position after standardization by calculating the proportional relationship between each point and the boundary. The final standard angle image is the result of all points after this transformation.

[0035] In an optional embodiment, clustering the electricity consumption sequence and water consumption sequence of each user includes: Set the number of clusters for the electricity consumption series and water consumption series, and randomly select a sample point as the initial cluster center; Calculate the shortest distance from each sample point to the existing cluster center, calculate the probability of each sample point becoming the next cluster center, and take the sample point with the maximum probability as the next cluster center; Repeat the above operation until a preset number of initial cluster centers are selected; Calculate the Euclidean distance between the sample point and the cluster center, compare the Euclidean distance between each sample point and each cluster center, and divide each sample point into the class where the nearest cluster center is located; Calculate the sample mean points of each category, update the cluster center, and repeat the above operations until the position of each cluster center no longer changes.

[0036] In an optional embodiment, the calculation formula for the energy-saving capability of users with abnormal electricity and water consumption is as follows: ; in, Indicates the energy-saving capabilities of users with abnormal electricity and water usage, Indicates Classification of electricity usage conditions and various water usage conditions Sampling period The electricity and water consumption at each moment, Indicates Types of electricity and water usage The electricity and water usage standards at the moment, Indicates the total types of electrical conditions and various water-using conditions. Indicates the total number of moments in the sampling period, Indicates the total number of sampling periods.

[0037] The embodiments of the present invention are described in detail above with reference to the accompanying drawings, but the present invention is not limited thereto, and various changes can be made within the knowledge scope of technicians in the relevant technical field without departing from the purpose of the present invention.

Claims

1. An energy consumption management system for urban residential areas based on the Internet of Things, characterized in that: It includes an image acquisition unit (1), an energy consumption recognition unit (2), an energy consumption statistics unit (3), a cost comparison unit (4), an abnormality detection unit (5), a data display unit (6) and an energy saving prediction unit (7), specifically: The image acquisition unit (1) is used to acquire images of electric meters and water meters in a target residential area based on a camera according to a preset sampling period, so as to obtain electric meter images and water meter images of each user, and transmit the electric meter images and water meter images to the energy consumption identification unit (2); The energy consumption recognition unit (2) is used to receive the electric meter image and the water meter image transmitted by the image acquisition unit (1), and perform image recognition on the electric meter image and the water meter image based on a pre-trained image recognition model to obtain the cumulative power consumption corresponding to the electric meter image and the cumulative water consumption corresponding to the water meter image, and obtain the current power consumption and the current water consumption based on the cumulative power consumption and the cumulative water consumption in the current acquisition cycle and the cumulative power consumption and the cumulative water consumption in the previous acquisition cycle, that is, obtain the power consumption and the water consumption of each user in the target residential area, and transmit the power consumption and the water consumption of each user to the energy consumption statistics unit (3) and the cost comparison unit (4).

2. According to the Internet of Things-based urban residential area energy consumption management system of claim 1, it is characterized in that: The fee comparison unit (4) is used to receive the electricity consumption and water consumption of each user transmitted by the energy consumption identification unit (2), and obtain the electricity and water charges of each user based on the preset electricity and water unit prices, obtain the prepaid electricity and water charges of each user based on the electricity and water charges of each user, compare the prepaid electricity and water charges of each user with the pre-acquired power supply bill and water supply bill to obtain the difference in electricity and water charges of each user, and make additional charges or refunds to each user based on the difference in electricity and water charges of each user; The energy consumption statistics unit (3) is used to receive the power consumption and water consumption of each user transmitted by the energy consumption identification unit (2), and to sort the power consumption and water consumption of each user according to a preset collection period to obtain the power consumption sequence and water consumption sequence of each user, and to obtain the power consumption sequence and water consumption sequence of each area in the target residential community based on the power consumption sequence and water consumption sequence of each user, and to transmit the power consumption sequence and water consumption sequence of each user to the abnormality detection unit (5), and to transmit the power consumption sequence and water consumption sequence of each area in the target residential community to the data display unit (6).

3. The urban residential area energy consumption management system based on the Internet of Things according to claim 2 is characterized in that: The abnormality detection unit (5) is used to receive the electricity consumption sequence and water consumption sequence of each user transmitted by the energy consumption statistics unit (3), and cluster the electricity consumption sequence and water consumption sequence of each user to obtain various electricity consumption conditions and various water consumption conditions, and cluster the electricity consumption sequence and water consumption sequence of each user again to determine the electricity consumption benchmarks of various electricity consumption conditions and the water consumption benchmarks of various water consumption conditions, and judge whether the electricity consumption sequence and water consumption sequence of each user are abnormal in electricity and water consumption based on the electricity consumption benchmarks of various electricity consumption conditions and the water consumption benchmarks of various water consumption conditions. If the electricity consumption and water consumption of each user are abnormal, the various electricity consumption conditions and various water consumption conditions as well as the electricity consumption benchmarks of various electricity consumption conditions and the water consumption benchmarks of various water consumption conditions are transmitted to the energy-saving prediction unit (7).

4. The urban residential area energy consumption management system based on the Internet of Things according to claim 3 is characterized in that: The energy-saving prediction unit (7) is used to receive various power consumption conditions and various water consumption conditions as well as power consumption benchmarks of various power consumption conditions and water consumption benchmarks of various water consumption conditions transmitted by the abnormality detection unit (5), and predict the energy-saving capabilities of users with abnormal power and water consumption based on the various power consumption conditions and various water consumption conditions as well as the power consumption benchmarks of various power consumption conditions and water consumption benchmarks of various water consumption conditions, and transmit the energy-saving capabilities of users with abnormal power and water consumption to the data display unit (6); The data display unit (6) is used to receive the electricity consumption sequence and water consumption sequence of each area in the target residential community transmitted by the energy consumption statistics unit (3) and the energy-saving capabilities of users with abnormal electricity and water consumption transmitted by the energy-saving prediction unit (7), obtain energy-saving electricity consumption and energy-saving water consumption based on the energy-saving capabilities of users with abnormal electricity and water consumption, display the electricity consumption sequence and water consumption sequence of each area in the target residential community, and display the energy-saving electricity consumption and energy-saving water consumption of users with abnormal electricity and water consumption.

5. The urban residential area energy consumption management system based on the Internet of Things according to claim 1 is characterized in that: The image recognition model includes a regional positioning module, a background filtering module, a correction module and a digital recognition module, wherein the regional positioning module is used to perform regional positioning on the visible number areas of the electric meter image and the water meter image to obtain the visible number area image, the background filtering module is used to detect the visible number area image to filter the background information of the visible number area image, thereby obtaining the digital image of the visible number area image, the correction module is used to correct the digital image to obtain a digital image at a standard angle, and the digital recognition module is used to perform digital recognition on the digital image at a standard angle to obtain the number corresponding to the digital image at the standard angle, thereby obtaining the cumulative electricity consumption corresponding to the electric meter image and the cumulative water consumption corresponding to the water meter image.

6. The urban residential area energy consumption management system based on the Internet of Things according to claim 5 is characterized in that: The region positioning module adopts the YOLOv4-tiny model, the background filtering module adopts the improved DMPNet model, and the digital recognition module adopts a convolutional neural network.

7. The urban residential area energy consumption management system based on the Internet of Things according to claim 6 is characterized in that: The background filtering module is based on inserting two control points along the upper and lower boundaries of the display area image to improve the loss function of the DMPNet model, wherein the improved loss function is as follows: ； in, represents the position loss of 8 fitting points, and Represents the predicted value and true value of a certain point, represents the sum of the centrality scores of the 8 fitting points, represents the error between the predicted value and the true value at a certain point, and .

8. The urban residential area energy consumption management system based on the Internet of Things according to claim 7 is characterized in that: Correcting the digital image to obtain a digital image at a standard angle includes: Acquire a corner point set and a control point set of the digital image, and calculate a minimum convex hull of the corner point set and the control point set, wherein the minimum convex hull represents a minimum polygon enclosed by the corner point set and the control point set; Calculate the minimum circumscribed rectangle of the minimum convex hull, and obtain the width and height of the minimum circumscribed rectangle; Traversing each point in the digital image, obtaining a first distance between each point and the left border of the digital image, and calculating a ratio of the first distance to a width of the minimum circumscribed rectangle; The upper and lower boundary points of the digital image are determined based on the ratio of the first distance to the minimum circumscribed rectangle width, the corner point set and the control point set, wherein the expressions of the upper and lower boundaries are as follows: ； in, Indicates the upper and lower boundaries, and Represents the left and right corners, and represents the control point, Indicates the ratio of the first distance to the width of the minimum enclosing rectangle; Traversing each point in the digital image, obtaining a second distance between each point and the upper boundary of the digital image, and calculating a ratio of the second distance to the height of the minimum circumscribed rectangle; The corresponding position of each point in the digital image is calculated based on the ratio of the upper and lower boundary points of the digital image to the second distance and the height of the minimum circumscribed rectangle, thereby obtaining a digital image at a standard angle, wherein the expression of the corresponding position in the digital image is as follows: ； in, represents the corresponding position of each point in the digital image, and Represents the upper and lower boundary points of the digital image, Indicates the ratio of the second distance to the height of the minimum enclosing rectangle.

9. The urban residential area energy consumption management system based on the Internet of Things according to claim 3 is characterized in that: Clustering of electricity consumption series and water consumption series of each user, including: The number of clusters of the electricity consumption sequence and the water consumption sequence is set, and a sample point is randomly selected as the initial cluster center; Calculate the shortest distance from each sample point to the existing cluster center, calculate the probability of each sample point becoming the next cluster center, and take the sample point with the maximum probability as the next cluster center; Repeat the above operation until a preset number of initial cluster centers are selected; Calculate the Euclidean distance between the sample point and the cluster center, compare the Euclidean distance between each sample point and each cluster center, and divide each sample point into the class where the nearest cluster center is located; Calculate the sample mean points of each category, update the cluster center, and repeat the above operations until the position of each cluster center no longer changes.

10. The urban residential area energy consumption management system based on the Internet of Things according to claim 1 is characterized in that: The calculation formula for the energy-saving capacity of users with abnormal electricity and water consumption is as follows: ； in, Indicates the energy-saving capabilities of users with abnormal electricity and water usage, Indicates Classification of electricity usage conditions and various water usage conditions Sampling period The electricity and water consumption at each moment, Indicates Types of electricity and water usage The electricity and water usage standards at the moment, Indicates the total types of electrical conditions and various water-using conditions. Indicates the total number of moments in the sampling period, Indicates the total number of sampling periods.

Images