Smart city monitoring system and method

Abstract

The invention discloses a smart city monitoring system and method, and relates to the field of city monitoring, and the system comprises a digital twin model, an intelligent cooperation module, a decision optimization module, an execution module, a feedback and adjustment module, an interactive display module, and a public participation platform. Meanwhile, the city modeling and the city entity are subjected to state synchronization work through the Internet of Things; the intelligent cooperation module is used for subdividing city monitoring into a plurality of sub-tasks, and each sub-task is responsible for a specific intelligent agent; and the decision optimization module performs decision optimization on the intelligent cooperation module through a reinforcement learning algorithm. According to the smart city monitoring system and method, omnibearing and three-dimensional city information collection is achieved, the coverage range is wide, data sources are rich and complementary, the city operation state can be obtained in real time, and sufficient data support can be provided for subsequent analysis and decision making.

Description

Technical Field

[0001] The present invention relates to the field of urban monitoring, and particularly to a smart city monitoring system and method. Background Art

[0002] In the current rapid urbanization process, the scale of cities is constantly expanding, the population is continuously growing, and the complexity of cities is increasing day by day, which also makes the importance of urban monitoring more prominent.

[0003] When monitoring cities, most rely on single or several of the above monitoring means, such as only relying on traffic cameras to monitor traffic and weather stations to monitor the environment, resulting in limited monitoring scope and single information of cities, making it difficult to comprehensively cover every corner of the city. At the same time, when conducting urban management based on urban monitoring, it mostly relies on fixed rules and simple data analysis, lacking adaptability and learning ability, resulting in problems such as lagging decisions and poor flexibility.

[0004] Therefore, it is necessary to propose a smart city monitoring system and method to solve the above problems. Summary of the Invention

[0005] The main purpose of the present invention is to provide a smart city monitoring system and method, which can effectively solve the problems in the background art.

[0006] To achieve the above purpose, the technical solution adopted by the present invention is as follows:

[0007] A smart city monitoring system includes a digital twin model, an intelligent collaboration module, a decision optimization module, an execution module, a feedback and adjustment module, an interactive display module, and a public participation platform. The digital twin model is used to model the city, and at the same time, through the Internet of Things, the state synchronization of the city model and the city entity is carried out.

[0008] The intelligent collaboration module is used to divide urban monitoring into multiple subtasks, and each subtask is responsible for a specific intelligent agent.

[0009] The decision optimization module optimizes the decision of the intelligent collaboration module through a reinforcement learning algorithm.

[0010] The execution module realizes the precise control of various urban devices and systems based on the analysis results of urban monitoring of the digital twin model and the decisions of multiple intelligent agents.

[0011] The feedback and adjustment module is used to collect actual effect data through the digital twin model after the execution module implements the decision, and then feedback it to the dynamic modeling module, the real-time state synchronization module, the intelligent collaboration module, and the decision optimization module through the Internet of Things.

[0012] Preferably, the digital twin model includes a dynamic modeling module and a real-time status synchronization module. The dynamic modeling module generates a three-dimensional terrain model including urban road slopes and building heights through high-precision lidar scanning; provides large-area and periodic urban images through satellite remote sensing for monitoring urban land use changes and vegetation cover dynamics; and obtains high-resolution texture images and three-dimensional point cloud data by low-altitude mapping of key urban areas and complex scenes by drones.

[0013] The digital twin model integrates building information by deeply combining building information model technology to present the internal structure and facilities of buildings.

[0014] Preferably, the real-time status synchronization module builds a two-way data transmission between urban entities and the digital twin model through the Internet of Things. By installing smart electricity meters, water meters, and gas meters in the city, it is used to collect energy consumption data in real time; installs environmental monitoring equipment to monitor air quality and noise levels; installs traffic cameras and geomagnetic sensors to collect traffic flow and vehicle speeds, and the above data is uploaded to the digital twin model through the Internet of Things.

[0015] The digital twin model sends instructions generated based on decisions to corresponding execution devices in the city through a reverse communication link to achieve precise control of urban entities.

[0016] Preferably, the agents are distributed in every corner of the city, covering the fields of transportation, energy, environment, and safety; the agents share and record data through distributed ledger technology to ensure the consistency and immutability of information. When a communication interruption occurs to a certain urban device due to an emergency, other agents can automatically adjust their cooperation strategies and transmit information through detour paths to maintain the coherence of urban monitoring.

[0017] Preferably, the decision optimization module is electrically connected to the intelligent cooperation model. During the process of the agents performing tasks, they select an action based on the current monitored urban status. After executing the action, they will receive a reward signal, enabling the agents to learn how to select the optimal action to maximize the long-term cumulative reward; the agents can share experiences with each other.

[0018] Preferably, the execution module uses edge computing devices to process some control decisions at edge nodes close to the devices to reduce data transmission latency and improve response speed.

[0019] Preferably, the feedback and adjustment module uses big data analysis and machine learning algorithms to evaluate the feedback data and judge the effectiveness of the decision. When it is found that there is a deviation between the actual effect and the expected goal, the analysis and decision-making process will be restarted. The digital twin model will simulate different decision-making schemes again, and the agents will also cooperate again to adjust the decision-making strategy to ensure the continuous optimization of the decision. The specific steps are as follows:

[0020] Collect and integrate the data of urban monitoring through agents, and clean and preprocess the data;

[0021] Extract and select the features valuable for decision evaluation from the collected and integrated raw data. Through the feature selection algorithm, screen out the features most relevant to the decision effect, remove redundant features, improve the training efficiency and accuracy of the model. The feature selection algorithm selects analysis of variance, divides the data according to different categories of decision effects, calculates the mean and variance of each feature under different categories, and calculates based on this

[0022] The F value, and the formula is:

[0023]

[0024] where MS between is the mean square between groups; MS within is the mean square within groups;

[0025] Select a suitable machine learning model according to the decision type and data characteristics. When evaluating traffic guidance decisions, a regression model is selected to predict the change in traffic flow after taking guidance measures; when judging the effectiveness of environmental protection policy decisions, a classification model is used to judge whether the environmental quality has improved, and the model is trained using historical data and the model parameters are adjusted to accurately capture the laws and patterns in the data;

[0026] Set corresponding evaluation indicators for different decision-making goals. When evaluating traffic decisions, indicators such as the average vehicle speed increase rate and congestion mitigation time are used to measure; when evaluating environmental protection decisions, indicators such as the pollutant concentration decrease ratio and the increase ratio of air quality compliance days are used, and these indicators can intuitively reflect the effect after the implementation of the decision;

[0027] Input the real-time feedback data into the trained model, the model outputs the prediction result, compare the prediction result with the actual observation value, and judge the effectiveness of the decision in combination with the evaluation indicators.

[0028] Preferably, the interactive display module provides urban monitoring for managers based on VR, AR, and MR. In the VR mode, view the operation status of the city from the first-person perspective, including monitoring the real-time situation of traffic congestion points and observing the operation status of internal facilities of buildings; in the AR and MR modes, virtual information can be combined with the real scene.

[0029] Preferably, the public participation platform is used to encourage urban citizens to participate in urban monitoring and management. The public participation platform enables citizens to report problems around them, including road damage, street lamp failures, and environmental pollution. The public participation platform includes a positioning module and a camera reporting module, and the positioning module and the camera reporting module are used to obtain the location where the problem occurs and on-site photos.

[0030] A smart city monitoring method includes the following operating steps:

[0031] S1: Perception construction, including low-altitude drone inspections, ground Internet of Things sensors, vehicle Internet of Things device linkages, and mobile terminal public reporting. Among them, low-altitude drone inspections include using drones for rapid monitoring of large urban areas;

[0032] Ground Internet of Things sensors are used to deploy geomagnetic sensors every 50 - 100 meters on urban roads for real-time monitoring of vehicle flow, vehicle speed, and vehicle types for traffic flow analysis and congestion prediction; at key parts of bridges, such as bridge piers and bridge body connections, stress sensors and displacement sensors are installed to collect data every 5 minutes for monitoring the structural health of the bridge; inside buildings, temperature and humidity sensors and air quality sensors are installed to adjust the indoor environment in real time to ensure the comfort and health of residents;

[0033] Vehicle Internet of Things device linkages are used to cooperate with operating vehicles such as taxis, buses, and logistics vehicles. Vehicle Internet of Things terminals are installed on the vehicles to upload the vehicle's location, driving speed, and driving route in real time. Combining traffic big data analysis, traffic congestion sections and times are predicted, and traffic warnings are issued in advance to guide vehicles to detour. At the same time, cameras on the vehicles are used to monitor the road conditions in real time;

[0034] Mobile terminal public reporting is used to develop a dedicated urban monitoring APP. Citizens can take photos and videos at any time and place through their mobile phones to report abnormal situations in the city, and certain integral rewards are given to citizens. The integral can be exchanged for daily necessities or participate in lucky draws to improve citizens' participation enthusiasm;

[0035] S2: Deep reinforcement learning analysis, including distributed agent collaboration and deep reinforcement learning decision-making. Among them, distributed agent collaboration is used to divide urban monitoring tasks into multiple subtasks, including traffic monitoring, environmental monitoring, energy monitoring, and public safety monitoring. Each subtask is responsible for by the corresponding agent; the agent ensures the consistency and security of data through distributed ledger technology;

[0036] Deep reinforcement learning decision-making, including each agent using the deep Q-network algorithm. The agent learns the optimal decision-making strategy based on continuously trying different actions, and at the same time stores successful experiences in the shared experience pool, enabling other agents to learn from them and accelerating the overall learning process;

[0037] S3: Decision execution, including intelligent execution, blockchain recording and traceability, and real-time feedback and adjustment. Among them, intelligent execution is based on the analysis of the digital twin model and the decision-making results of the agent to achieve precise control of various urban devices;

[0038] Blockchain recording and traceability is used to record key data in the decision execution process, such as decision time, execution device, and execution result, on the blockchain. When problems occur, the entire process of decision execution can be quickly queried to clarify the responsible entity;

[0039] Real-time feedback and adjustment. After the decision is executed, data on the execution effect is collected in real time through deployed sensors, and big data analysis tools and machine learning algorithms are used to evaluate the data. When it is found that there is a deviation between the actual effect and the expected goal, the digital twin model re-simulates different decision-making schemes, and the agents re-cooperate to adjust the decision-making strategy until the dynamic optimization of the decision is achieved.

[0040] Compared with the prior art, the present invention provides a smart city monitoring system and method, which have the following beneficial effects:

[0041] 1. The smart city monitoring system and method achieve all-round and three-dimensional urban information collection, with a wide coverage range and rich and complementary data sources. It can obtain the operating status of the city in real time, including multi-faceted information on traffic, environment, and buildings, providing sufficient data support for subsequent analysis and decision-making. At the same time, it can encourage public participation, enhancing citizens' sense of responsibility and participation in urban monitoring management, and solving the problems of limited monitoring range and single information when monitoring the city. This system can discover problems such as road diseases and missing manhole covers in the city through public reporting and drone inspections, and at the same time can achieve real-time monitoring of a wider area.

[0042] 2. The smart city monitoring system and method make the system highly flexible and autonomous through distributed agent cooperation. Each agent can make quick decisions according to local conditions, and can also optimize the overall decision through shared experiences. Deep reinforcement learning enables agents to continuously learn and accumulate, adapt to the complex and changing urban environment, and make better decisions, solving the problems of lagging decision-making and lack of flexibility in urban monitoring.

[0043] 3. When making decisions and implementing them, the smart city monitoring system and method have a transparent and traceable process, with secure and reliable data, ensuring the accuracy of decision implementation and clear responsibility. The real-time feedback and adjustment mechanism enables the system to optimize decisions in a timely manner according to the actual implementation effects, improving the efficiency and quality of urban management, solving the problems of opaque decision implementation processes and difficult traceability of responsibilities in case of problems. It can use blockchain traceability for urban energy distribution decision-making and implementation to ensure fairness and justice, and quickly locate problems in case of anomalies.

[0044] 4. Through the public participation platform, the smart city monitoring system and method enable citizens to actively participate in urban monitoring and management. They can not only solve practical problems around them but also stimulate citizens' innovative thinking, and receive innovative suggestions from citizens regarding urban space utilization and environmental protection and energy conservation. These suggestions, after evaluation and implementation, can bring new ideas and vitality to urban development. At the same time, the interactive platform allows urban managers to conduct urban monitoring and decision-making simulations in a virtual environment, enabling them to discover some potential problems in advance and avoid unnecessary losses during the actual implementation process. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] Figure 1 It is the system block diagram of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0046] To make the technical means, creative features, achieved objectives, and functions of the present invention easy to understand, the present invention will be further described below in conjunction with specific embodiments.

[0047] Embodiment 1:

[0048] As Figure 1 shown, a smart city monitoring system includes a digital twin model, an intelligent collaboration module, a decision optimization module, an execution module, a feedback and adjustment module, an interactive display module, and a public participation platform. The digital twin model is based on modeling the city, and at the same time synchronizes the states of the city modeling and the urban entity through the Internet of Things;

[0049] The digital twin model includes a dynamic modeling module and a real-time status synchronization module. The dynamic modeling module generates a three-dimensional terrain model containing the slope of urban roads and building heights through high-precision lidar scanning; provides large-area and periodic urban images through satellite remote sensing for monitoring urban land use changes and vegetation cover dynamics; obtains high-resolution texture images and three-dimensional point cloud data by low-altitude mapping of key urban areas and complex scenes using drones;

[0050] The digital twin model integrates building information by deeply combining with Building Information Modeling (BIM) technology, and is used to present the internal structure and facilities of a building. In intelligent building monitoring, it can monitor the temperature, humidity, lighting and equipment operation status of each floor in real time. For urban dynamic elements, with the help of the Internet of Vehicles (IoV) technology, vehicle-mounted sensors communicate with roadside base stations to obtain the position, speed and driving direction of vehicles in real time, which is used to accurately track the trajectory of each vehicle. Mobile signaling data is utilized to analyze the large-scale movement characteristics of people, such as the crowd gathering and evacuation directions during the morning and evening rush hours on weekdays. At the same time, cameras deployed in public places use computer vision technology to identify the behaviors and activities of pedestrians.

[0051] The real-time status synchronization module builds a two-way data transmission between urban entities and the digital twin model through the Internet of Things. By installing smart electricity meters, water meters and gas meters in the city, it is used to collect energy consumption data in real time; installing environmental monitoring equipment to monitor air quality and noise levels; installing traffic cameras and geomagnetic sensors to collect traffic flow and vehicle speed. The above data is uploaded to the digital twin model through the Internet of Things.

[0052] Based on the instructions generated by the decision-making, the digital twin model sends them to the corresponding execution devices in the city through the reverse communication link, which is used to achieve precise control of urban entities. When there is heavy rain and waterlogging in a certain area of the city, the water level sensors deployed on the road upload the water depth data to the digital twin model in real time. The model quickly analyzes the impact of the waterlogging on the surrounding traffic and drainage systems, and generates countermeasures, such as adjusting the duration of traffic lights nearby to guide vehicles to detour, and starting drainage pumping stations to increase the drainage volume. The relevant instructions are sent to the traffic light control system and drainage pumping stations through the communication network to promptly relieve the waterlogging problem and reduce the impact on urban operations.

[0053] The intelligent collaboration module is used to divide urban monitoring into multiple subtasks, and each subtask is responsible for a specific intelligent agent.

[0054] Intelligent agents are distributed in every corner of the city, covering the fields of transportation, energy, environment and security. In the transportation field, the traffic light control intelligent agents at each intersection collect the surrounding traffic flow and vehicle queue length information, and interact with the intelligent agents at adjacent intersections. Data sharing and recording are carried out among intelligent agents through distributed ledger technology to ensure the consistency and immutability of information. When a communication interruption occurs to a certain urban device due to an emergency, other intelligent agents can automatically adjust the collaboration strategy and transmit information through detour paths to maintain the coherence of urban monitoring.

[0055] The decision optimization module optimizes the decisions of the intelligent collaboration module through reinforcement learning algorithms.

[0056] The decision optimization module is electrically connected to the intelligent cooperation model. During the task execution process, the intelligent agent selects an action based on the currently monitored urban status. After executing the action, it will receive a reward signal, which is used for the intelligent agent to learn how to select the optimal action to maximize the long-term cumulative reward. Intelligent agents can share experiences. After an intelligent agent in a certain area successfully responds to a pollution event, it records its decision-making process and experiences in the shared experience pool. When other intelligent agents encounter similar situations, they can draw on these experiences to find the optimal solution more quickly. In urban energy monitoring, the energy distribution intelligent agent decides the amount of electricity to be distributed to different regions based on the electricity demand, power generation cost, and renewable energy generation situation at different times. When, during a certain period, the energy cost is reduced and user needs are met through optimized distribution strategies, the intelligent agent will receive a reward. After multiple learning processes, the intelligent agent can make the optimal energy distribution decision under different energy supply and demand scenarios, achieving the efficient utilization of urban energy.

[0057] The execution module realizes the precise control of various urban devices and systems based on the analysis results of the digital twin model for urban monitoring and the decisions of multiple intelligent agents.

[0058] The execution module uses edge computing devices to process some control decisions at the edge nodes close to the devices, aiming to reduce data transmission latency and improve response speed. In industrial parks, the edge computing devices continuously monitor the operating parameters of production equipment. When abnormalities are detected, alarms are immediately issued and corresponding control measures are taken to ensure production safety. By ensuring the monitored content, the decisions of urban monitoring and management can be accurately and timely implemented in the actual environment. Through the refined control of various devices, the urban resource utilization efficiency is improved, the operation cost is reduced, and the urban service quality is enhanced. In the field of intelligent transportation, the traffic signal control system changes the duration and phase of traffic lights in real time according to the signal light adjustment strategy analyzed by multiple intelligent agents to optimize the traffic flow. In the urban lighting system, the intelligent street lamp controller automatically adjusts the brightness of street lamps according to the environmental brightness and the flow of pedestrians and vehicles, achieving a balance between energy conservation and lighting requirements.

[0059] The feedback and adjustment module is used to collect actual effect data through the digital twin model after the execution module implements the decision, and then feedback it to the dynamic modeling module, real-time status synchronization module, intelligent cooperation module, and decision optimization module through the Internet of Things.

[0060] The feedback and adjustment module uses big data analysis and machine learning algorithms to evaluate the feedback data and judge the effectiveness of the decision. When it is found that there is a deviation between the actual effect and the expected goal, the analysis and decision-making process will be restarted. The digital twin model will simulate different decision-making schemes again, and the agents will also cooperate again to adjust the decision-making strategy to ensure the continuous optimization of the decision. The feedback and adjustment module is used to make the city have self-adaptability and self-optimization ability when monitoring the city, and can adjust the management strategy in a timely manner according to the real-time changes of the city to avoid urban operation problems caused by fixed rules or lagged decisions. The specific steps are as follows:

[0061] Collect and integrate the data monitored by the agents for the city, and clean and preprocess the data;

[0062] Extract and select the features valuable for decision evaluation from the collected and integrated raw data. In traffic decision evaluation, not only consider the traffic flow and vehicle speed, but also calculate derivative features such as traffic congestion index and vehicle queue length. Through the feature selection algorithm, screen out the features most relevant to the decision effect, remove redundant features, improve the model training efficiency and accuracy. The feature selection algorithm selects analysis of variance, divides the data according to different categories of decision effects, calculates the mean and variance of each feature under different categories, and calculates the F value based on this. The formula is:

[0063]

[0064] where MS between is the mean square between groups; MS within is the mean square within groups;

[0065] Select a suitable machine learning model according to the decision type and data characteristics. When evaluating traffic guidance decisions, select a regression model to predict the change in traffic flow after taking guidance measures; when judging the effectiveness of environmental protection policy decisions, use a classification model to judge whether the environmental quality has improved, train the model with historical data, and adjust the model parameters to accurately capture the laws and patterns in the data;

[0066] Set corresponding evaluation indicators for different decision-making goals. When evaluating traffic decisions, use indicators such as average vehicle speed increase rate and congestion relief time to measure; when evaluating environmental protection decisions, use indicators such as pollutant concentration decrease ratio and increased proportion of days with air quality meeting standards. These indicators can intuitively reflect the effect after the implementation of the decision;

[0067] Input the real-time feedback data into the trained model, the model outputs the prediction result, compare the prediction result with the actual observation value, and judge the decision effectiveness in combination with the evaluation indicators. If after the implementation of the traffic guidance decision, the model predicts that the average vehicle speed increases by 15%, and the actual increase is 12%, which is close to the predicted value, it indicates that the decision is effective; if the deviation is too large, the reason needs to be analyzed and the decision adjusted.

[0068] The interactive display module provides urban monitoring for managers based on VR, AR, and MR. In the VR mode, the operation status of the city can be viewed from the first-person perspective, including monitoring the real-time situation of traffic congestion points and observing the operation status of internal facilities of buildings. In the AR and MR modes, virtual information can be combined with the real scene. When on-site inspecting urban construction projects, virtual information such as the project planning model, progress information, and future effects can be superimposed on the real scene through mobile phones and smart glasses for intuitive understanding of the project situation.

[0069] The public participation platform is used to encourage urban citizens to participate in urban monitoring and management. The public participation platform enables citizens to report problems around them, including road damage, street lamp failures, and environmental pollution. The public participation platform includes a positioning module and a camera reporting module, which are used to obtain the location where the problem occurs and on-site photos.

[0070] Embodiment 2:

[0071] A smart city monitoring method includes the following operating steps:

[0072] S1: Perception construction, including low-altitude drone inspections, ground Internet of Things sensors, vehicle Internet of Things device linkage, and public reporting by mobile terminals. Among them, low-altitude drone inspections include using drones for rapid monitoring of large areas of the city;

[0073] Ground Internet of Things sensors are used to deploy geomagnetic sensors every 50 - 100 meters on urban roads for real-time monitoring of vehicle flow, vehicle speed, and vehicle types for traffic flow analysis and congestion prediction; at key parts of bridges, such as bridge piers and bridge body joints, stress sensors and displacement sensors are installed to collect data every 5 minutes for monitoring the structural health of the bridge; inside buildings, temperature and humidity sensors and air quality sensors are installed to adjust the indoor environment in real time to ensure the comfort and health of residents;

[0074] Vehicle Internet of Things device linkage is used to cooperate with operating vehicles such as taxis, buses, and logistics vehicles. Vehicle Internet of Things terminals are installed on the vehicles to upload the vehicle location, driving speed, and driving route in real time. Combining traffic big data analysis, traffic congestion sections and times are predicted, and traffic warnings are issued in advance to guide vehicles to detour. At the same time, the cameras on the vehicles are used to monitor the road conditions in real time;

[0075] Public reporting by mobile terminals is used to develop a dedicated urban monitoring APP. Citizens can take photos and videos at any time and place through their mobile phones to report abnormal situations in the city, and certain integral rewards are given to citizens. The integral can be exchanged for daily necessities or participate in lucky draws to improve citizens' participation enthusiasm;

[0076] S2: Deep reinforcement learning analysis, including distributed agent collaboration and deep reinforcement learning decision-making. Among them, distributed agent collaboration is used to divide urban monitoring tasks into multiple subtasks, including traffic monitoring, environmental monitoring, energy monitoring, and public security monitoring. Each subtask is responsible for by the corresponding agent; the agent ensures the consistency and security of data through distributed ledger technology;

[0077] Deep reinforcement learning decision-making, including that each agent uses the deep Q-network algorithm. The agent learns the optimal decision-making strategy based on continuously trying different actions, and at the same time stores the successful experience in the shared experience pool, so that other agents can learn from it and accelerate the overall learning process;

[0078] S3: Decision execution, including intelligent execution, blockchain recording and traceability, and real-time feedback and adjustment. Among them, intelligent execution is based on the analysis of the digital twin model and the decision-making results of the agent, and realizes the precise control of various urban devices. In the intelligent lighting system, according to information such as environmental brightness, pedestrian and vehicle flow, the edge computing device processes data in real time near the street lamp and automatically adjusts the street lamp brightness. In the intelligent transportation system, the traffic signal controller adjusts the signal light duration and phase in real time according to the swarm intelligence decision-making;

[0079] Blockchain recording and traceability is used to record key data in the decision execution process, such as decision time, execution device, and execution result on the blockchain. When a problem occurs, the whole process of decision execution can be quickly queried to clarify the responsible entity;

[0080] Real-time feedback and adjustment. After the decision is executed, the execution effect data is collected in real time through the deployed sensors, and the data is evaluated using big data analysis tools and machine learning algorithms. When it is found that there is a deviation between the actual effect and the expected goal, the digital twin model re-simulates different decision-making schemes, the agents re-collaborate, and adjust the decision-making strategy until the dynamic optimization of the decision is achieved.

[0081] The above shows and describes the basic principles, main features and advantages of the present invention. Those skilled in the art of this industry should understand that the present invention is not limited by the above embodiments. What is described in the above embodiments and the specification only illustrates the principles of the present invention. Without departing from the spirit and scope of the present invention, the present invention will have various changes and improvements, and these changes and improvements all fall within the scope of the present invention claimed. The scope of protection claimed by the present invention is defined by the appended claims and their equivalents.

Claims

1. A smart city monitoring system, including a digital twin model, an intelligent collaboration module, a decision optimization module, an execution module, a feedback and adjustment module, an interactive display module, and a public participation platform, characterized in that: The digital twin model is based on modeling the city, and synchronizes the status of the city modeling and city entities through the Internet of Things; The intelligent collaboration module is used to subdivide urban monitoring into multiple subtasks, each of which is responsible for a specific intelligent agent; The decision optimization module optimizes the decision of the intelligent collaboration module through a reinforcement learning algorithm; The execution module realizes precise control of various equipment and systems in the city based on the analysis results of the digital twin model on urban monitoring and the decision-making of multiple intelligent agents; The feedback and adjustment module is used to collect actual effect data through the digital twin model after the execution module implements the decision, and then feed back to the dynamic modeling module, real-time status synchronization module, intelligent collaboration module, and decision optimization module through the Internet of Things.

2. A smart city monitoring system according to claim 1, characterized in that: The digital twin model includes a dynamic modeling module and a real-time state synchronization module. The dynamic modeling module generates a three-dimensional terrain model including urban road slopes and building heights through high-precision laser radar scanning; provides large-area, periodic urban images through satellite remote sensing to monitor urban land use changes and vegetation coverage dynamics; and obtains high-resolution texture images and three-dimensional point cloud data through low-altitude mapping of key urban areas and complex scenes by drones. The digital twin model integrates building information by deeply combining building information modeling technology to present the internal structure and facilities of the building.

3. A smart city monitoring system according to claim 2, characterized in that: The real-time status synchronization module builds two-way data transmission between the city entity and the digital twin model through the Internet of Things. It collects energy consumption data in real time by installing smart electricity meters, water meters, and gas meters in the city; installs environmental monitoring equipment to monitor air quality and noise levels; installs traffic cameras and geomagnetic sensors to collect traffic flow and vehicle speed. The above data is uploaded to the digital twin model through the Internet of Things; The digital twin model generates instructions based on decisions and sends them to corresponding execution devices in the city through a reverse communication link to achieve precise control of urban entities.

4. A smart city monitoring system according to claim 1, characterized in that: The intelligent agents are distributed in every corner of the city, covering the fields of transportation, energy, environment, and security. The intelligent agents share and record data through distributed ledger technology to ensure the consistency and non-tamperability of information. When a city device loses communication due to an emergency, other intelligent agents can automatically adjust their collaboration strategies and transmit information through circuitous paths to maintain the continuity of city monitoring.

5. A smart city monitoring system according to claim 1, characterized in that: The decision optimization module is electrically connected to the intelligent collaboration model. When executing a task, the intelligent agent selects an action based on the currently monitored city status. After executing the action, it will receive a reward signal for the intelligent agent to learn how to choose the optimal action to maximize the long-term cumulative reward. The intelligent agents can share experience.

6. A smart city monitoring system according to claim 1, characterized in that: The execution module uses edge computing devices to process some control decisions at edge nodes close to the devices, so as to reduce data transmission delays and improve response speed.

7. A smart city monitoring system according to claim 1, characterized in that: The feedback and adjustment module uses big data analysis and machine learning algorithms to evaluate the feedback data and judge the effectiveness of the decision. When it is found that the actual effect deviates from the expected goal, the analysis and decision-making process is restarted. The digital twin model will simulate different decision-making plans again, and the intelligent agents will collaborate again to adjust the decision-making strategy to ensure the continuous optimization of the decision. Specifically, the following steps are included: Collect and integrate urban monitoring data through intelligent agents, and clean and pre-process the data; Extract and select features that are valuable for decision evaluation from the collected and integrated raw data. Through the feature selection algorithm, screen out the features most relevant to the decision effect, remove redundant features, and improve the efficiency and accuracy of model training. The feature selection algorithm uses variance analysis to divide the data according to different categories of decision effects, calculate the mean and variance of each feature in different categories, and calculate the F value based on this. The formula is: Among them, MS between is the mean square between groups; MS within is the within-group mean square; Select appropriate machine learning models based on decision types and data characteristics. When evaluating traffic diversion decisions, use regression models to predict changes in traffic volume after diversion measures are taken. When judging the effectiveness of environmental policy decisions, use classification models to determine whether environmental quality has improved. Use historical data to train the model and adjust model parameters to accurately capture the laws and patterns in the data. Corresponding evaluation indicators are set for different decision-making goals. When evaluating traffic decisions, indicators such as the average vehicle speed increase rate and congestion relief time are used; when evaluating environmental protection decisions, indicators such as the percentage of pollutant concentration reduction and the percentage of days with air quality reaching standards are used. These indicators can directly reflect the effect of the decision after implementation; The real-time feedback data is input into the trained model, and the model outputs the prediction results. The prediction results are compared with the actual observation values, and the effectiveness of the decision is judged based on the evaluation indicators.

8. A smart city monitoring system according to claim 1, characterized in that: The interactive display module provides city monitoring for managers based on VR, AR and MR. In VR mode, the operating status of the city can be viewed from a first-person perspective, including monitoring the real-time situation of traffic congestion points and observing the operating status of facilities inside buildings; in AR and MR modes, virtual information can be combined with real scenes.

9. A smart city monitoring system according to claim 1, characterized in that: The public participation platform is used to encourage urban citizens to participate in urban monitoring and management. The public participation platform allows citizens to report problems around them, including road damage, street light failure, and environmental pollution. The public participation platform includes a positioning module and a camera reporting module. The positioning module and the camera reporting module are used to obtain the location where the problem occurs and on-site photos.

10. A smart city monitoring method, using a smart city monitoring system as claimed in any one of claims 1 to 9, characterized in that: The steps include: S1: Perception construction, including low-altitude drone inspections, ground IoT sensors, vehicle networking equipment linkage, and mobile terminal public reporting. Low-altitude drone inspections include the use of drones to conduct rapid monitoring of large areas in the city; Ground IoT sensors are used to deploy geomagnetic sensors every 50-100 meters on urban roads to monitor vehicle flow, speed, and vehicle type in real time for traffic flow analysis and congestion prediction; Stress sensors and displacement sensors are installed at key parts of the bridge, such as piers and bridge body joints, to collect data every 5 minutes to monitor the health of the bridge structure; Install temperature and humidity sensors and air quality sensors inside buildings to adjust the indoor environment in real time to ensure the comfort and health of residents; The linkage of Internet of Vehicles equipment is used to cooperate with operating vehicles such as taxis, buses, and logistics vehicles. Internet of Vehicles terminals are installed on vehicles to upload vehicle location, driving speed, and driving route in real time. Combined with traffic big data analysis, traffic congestion sections and times are predicted, traffic warnings are issued in advance, and vehicles are guided to detour. At the same time, the cameras on the vehicles are used to monitor road conditions in real time; Mobile terminal public reporting is used to develop a special city monitoring APP. Citizens can take photos and videos anytime and anywhere through their mobile phones to report abnormal situations in the city. Citizens will be rewarded with certain points, which can be exchanged for daily necessities or participate in lucky draws to increase the enthusiasm of citizens to participate. S2: Deep reinforcement learning analysis, including distributed agent collaboration and deep reinforcement learning decision-making. Distributed agent collaboration is used to subdivide urban monitoring tasks into multiple subtasks, including traffic monitoring, environmental monitoring, energy monitoring, and public safety monitoring. Each subtask is managed by a corresponding agent. The agent ensures data consistency and security through distributed ledger technology. Deep reinforcement learning decision-making, including each intelligent agent using the deep Q network algorithm. The intelligent agent learns the optimal decision-making strategy by constantly trying different actions, and stores successful experiences in a shared experience pool so that other intelligent agents can learn from them and accelerate the overall learning process; S3: Decision execution, including intelligent execution, blockchain recording and tracing, and real-time feedback and adjustment. Intelligent execution is based on digital twin model analysis and the decision results of the intelligent agent to achieve precise control of various equipment in the city; Blockchain recording and tracing is used to record key data in the decision-making execution process, such as decision time, execution equipment, and execution results, on the blockchain. When problems arise, the entire process of decision execution can be quickly queried to identify the responsible party; Real-time feedback and adjustment: After a decision is executed, the execution effect data is collected in real time through the deployed sensors, and the data is evaluated using big data analysis tools and machine learning algorithms. When it is found that the actual effect deviates from the expected goal, the digital twin model re-simulates different decision-making plans, the intelligent agents re-collaborate, and adjust the decision-making strategy until dynamic optimization of the decision is achieved.

Images