A smart campus big data cloud edge collaborative collection management method and system

Abstract

This invention provides a cloud-edge collaborative data collection and management method and system for smart campuses. It collects multi-source heterogeneous data from various scenarios within a smart campus, performs adaptive preprocessing and hierarchical transmission of the collected data at the edge, dynamically schedules cloud-edge collaborative tasks based on a digital twin virtual campus model, and employs a federated learning mechanism to perform global intelligent analysis while protecting data privacy. This generates early warning signals for the smart campus's operational status and drives the digital twin model for visualization. This invention effectively solves the problems of high bandwidth pressure, high latency, high privacy and security risks, and low resource utilization inherent in traditional centralized data processing models for smart campuses, achieving efficient, secure, and intelligent data collection and management for smart campuses.

Description

Technical Field

[0001] This invention relates to the field of smart campus technology, and in particular to a smart campus big data cloud-edge collaborative collection and management method and system. Background Technology

[0002] With the rapid development of information technology, the construction of smart campuses has become an important direction for educational informatization. Smart campuses deploy a large number of sensing devices to collect data from various scenarios within the campus, including teaching, security, energy consumption, logistics, and transportation. They then utilize big data and artificial intelligence technologies to achieve intelligent management and services. However, traditional smart campus data collection and management models primarily employ a centralized architecture, where all collected data is directly uploaded to cloud servers for processing and storage. This model suffers from the following problems: First, as the number of sensing devices increases, the amount of data generated grows exponentially, placing enormous pressure on network bandwidth and resulting in high data transmission latency, failing to meet the real-time requirements of business operations. Second, uploading large amounts of raw data to the cloud not only increases the computational and storage burden on cloud servers but also poses risks of data leakage and privacy security. Third, sensing devices in different scenarios use different communication protocols and data formats, creating data silos and making it difficult to achieve multi-source data fusion and analysis. Fourth, the lack of a dynamic task scheduling mechanism prevents the reasonable allocation of computing tasks based on the load of edge nodes and network conditions, leading to low resource utilization.

[0003] While some cloud-edge collaborative data processing solutions have emerged in existing technologies, most of these solutions are general-purpose and not optimized for the specific business characteristics of smart campuses. For example, existing solutions do not consider the different business priorities of various scenarios on campus, making it impossible to achieve hierarchical data transmission; they lack effective privacy protection technologies, making it easy to leak sensitive student and faculty data during global data analysis; and they do not incorporate digital twin technology, failing to achieve intuitive visualization of campus operational status and global collaborative management. Therefore, how to provide an efficient, secure, and intelligent cloud-edge collaborative data collection and management method and system for smart campuses has become an urgent technical problem to be solved. Summary of the Invention

[0004] Based on this, the present invention provides a smart campus big data cloud-edge collaborative collection and management method and system that can effectively solve the problems of high bandwidth pressure, high latency, high privacy and security risks, and low resource utilization in the traditional centralized data processing mode of smart campuses.

[0005] In a first aspect, the present invention provides a smart campus big data cloud-edge collaborative collection and management method, comprising the following steps: Collect multi-source heterogeneous data from various scenarios in smart campuses; Adaptive preprocessing and hierarchical transmission are performed on the edge side for the collected multi-source heterogeneous data; Dynamic scheduling of cloud-edge collaborative tasks based on a digital twin virtual campus model; A federated learning mechanism is used to perform global intelligent analysis while protecting data privacy. Generate early warning signals for the operation status of the smart campus and drive the digital twin model to visualize the display.

[0006] This invention constructs a cloud-edge-device collaborative smart campus big data collection and management architecture. First, it collects heterogeneous data from multiple sources across the campus using a global sensing network. Data processing and analysis tasks are deployed in layers, with basic preprocessing and hierarchical transmission completed at the edge. Based on a digital twin virtual campus model, global collaborative scheduling of cloud-edge resources and tasks is achieved. A federated learning mechanism is employed to complete global intelligent modeling without data leaving the local machine. Finally, the analysis results are transformed into early warning signals and visualized through the digital twin model. The entire process achieves a closed-loop data processing chain from collection, processing, and analysis to application, while simultaneously ensuring real-time performance, security, and resource utilization efficiency. This invention replaces the traditional centralized data processing model with a cloud-edge collaborative architecture, fundamentally solving problems such as high bandwidth pressure, high transmission latency, and significant privacy and security risks inherent in centralized architectures. Digital twin technology enables unified perception and collaborative management of the campus's overall status, while federated learning technology breaks down data silos while protecting data privacy, achieving fusion analysis of multi-source data. The overall solution significantly improves the efficiency and security of smart campus data processing, reduces system operating costs, and provides comprehensive and reliable data support and decision-making basis for intelligent campus management. As a further improvement to the technical solution of this invention, the collection of multi-source heterogeneous data in multiple scenarios of smart campuses specifically includes: Deploy various sensing devices in teaching, security, energy consumption, logistics, and transportation scenarios in smart campuses; The system collects structured, semi-structured, and unstructured data generated by various sensing devices through a unified edge data access gateway. Add a unique identifier, timestamp, and scene tag to each collected data.

[0007] This invention addresses the multi-scenario data collection needs of smart campuses by deploying compatible sensing devices in core scenarios such as teaching, security, energy consumption, logistics, and transportation. A unified edge data access gateway enables seamless access for sensing devices with different protocols and types. All collected data is automatically labeled with unique identifiers, timestamps, and scene tags, achieving standardized collection and unified management of multi-source heterogeneous data. This invention solves the problems of incompatible protocols and inconsistent data formats among sensing devices in different scenarios through a unified edge data access gateway, effectively breaking down data silos between various departments and systems within the campus. Adding unique identifiers, timestamps, and scene tags to each data point ensures data traceability and relevance, providing a comprehensive, accurate, and standardized data foundation for subsequent data processing, analysis, and application.

[0008] As a further improvement to the technical solution of this invention, the edge-side adaptive preprocessing and hierarchical transmission of the collected multi-source heterogeneous data specifically includes: The collected raw data is cleaned, denoised, format-converted, and feature-extracted. Based on preset business priority rules, the pre-processed data is divided into three levels: real-time critical data, important business data, and non-urgent statistical data. Real-time critical data is directly uploaded to the cloud server. Important business data is initially analyzed by edge nodes and the analysis results are uploaded. Non-urgent statistical data is stored locally on edge nodes and uploaded in batches periodically.

[0009] This invention decentralizes data preprocessing tasks to edge nodes, performing cleaning, noise reduction, format conversion, and feature extraction on the raw data at the edge. This removes invalid data and noise interference, converts data from different formats to a unified format, and extracts key features. Based on preset business priority rules, the preprocessed data is divided into three levels, and differentiated transmission strategies are adopted for each level, achieving adaptive hierarchical data transmission. This invention significantly reduces the amount of data that needs to be uploaded to the cloud through edge-side preprocessing, effectively reducing network bandwidth pressure and the computational and storage burden on cloud servers. Hierarchical transmission based on business priority ensures that critical business data with high real-time requirements is processed and responded to promptly, while also making full use of network idle periods to transmit non-urgent data, significantly improving network resource utilization and overall system efficiency.

[0010] As a further improvement to the technical solution of this invention, the dynamic scheduling of cloud-edge collaborative tasks based on a digital twin virtual campus model specifically includes: Construct a digital twin virtual campus model that corresponds one-to-one with the physical campus, and map the equipment status, environmental parameters and business operation status of the physical campus in real time; Real-time monitoring of computing resources, storage resources, network bandwidth and load of each edge node, and synchronization of this status information to the digital twin model; Based on the global state information in the digital twin model, a reinforcement learning algorithm is used to dynamically adjust the cloud-edge task allocation strategy, so as to reasonably allocate computing tasks to cloud or edge nodes for execution.

[0011] This invention constructs a digital twin virtual campus model that corresponds one-to-one with the physical campus, synchronizing the equipment status, environmental parameters, and business operation status of the physical campus in real time. It continuously monitors the computing resources, storage resources, network bandwidth, and load status of each edge node and maps this status information to the digital twin model in real time. Based on the global state view provided by the digital twin model, a reinforcement learning algorithm is used to dynamically adjust the cloud-edge task allocation strategy, rationally distributing computing tasks to cloud or edge nodes for execution according to the real-time status of each node. This invention achieves intuitive visualization and unified management of the overall campus operation status through digital twin technology, providing comprehensive and accurate global state information for cloud-edge collaborative task scheduling. The use of a reinforcement learning algorithm for dynamic task scheduling can adaptively adjust the task allocation scheme according to the real-time system status, effectively avoiding the problem of some nodes being overloaded while others are idle, significantly improving the system's resource utilization and overall response speed.

[0012] As a further improvement to the technical solution of this invention, the adoption of a federated learning mechanism for global intelligent analysis while protecting data privacy specifically includes: Train machine learning models for specific scenarios locally on each edge node; Each edge node encrypts the trained model parameters and uploads them to the cloud server; The cloud server aggregates the model parameters uploaded by all edge nodes to generate a global model; The updated global model parameters are distributed to each edge node to guide the iterative optimization of the local model on the edge nodes.

[0013] This invention employs a horizontal federated learning architecture, where each edge node trains a machine learning model for a specific scenario using its own stored data, with the original data remaining locally on each edge node throughout the training process. Each edge node only uploads the trained model parameters to the cloud server. The cloud server aggregates the model parameters uploaded by all edge nodes to generate a global model, and then distributes the updated global model parameters back to each edge node to guide the iterative optimization of the local model. This invention achieves joint modeling across multiple edge nodes without leaking the original data, effectively solving the data silo problem, fully utilizing the data resources of each edge node to improve model performance, and fundamentally avoiding the risk of leakage of sensitive student and faculty data during transmission and processing. Furthermore, the distributed training mode fully utilizes the computing resources of the edge nodes, further reducing the computational burden on the cloud server.

[0014] As a further improvement to the technical solution of this invention, generating a smart campus operation status early warning signal and driving the digital twin model for visualization specifically includes: Based on the results of global intelligent analysis, the operational status of various scenarios in the smart campus is monitored in real time; When an abnormal state or potential risk is detected, an early warning signal is generated that includes the location, type, severity, and recommended handling measures of the abnormality. The warning signal is synchronized to the digital twin virtual campus model, which drives the virtual entities in the corresponding areas of the model to perform visual alarms such as highlighting and flashing, and displays detailed warning information in the virtual interface.

[0015] This invention, based on global intelligent analysis results, monitors the operational status of various scenarios within a smart campus in real time, establishing a multi-dimensional anomaly detection model capable of identifying various abnormal states and potential risks. When an anomaly is detected, a structured early warning signal is automatically generated, containing the anomaly's location, type, severity, and suggested handling measures. This signal is synchronized to the digital twin virtual campus model, driving the corresponding virtual entities in the model to display visual alarms such as highlighting and flashing, while simultaneously displaying detailed warning information in the virtual interface. This invention can promptly detect various abnormal states and potential risks during campus operations, achieving a shift from passive response to proactive early warning. Visual alarms via the digital twin model can intuitively and accurately display the location and related information of anomalies, helping administrators quickly locate problems and take corresponding measures, significantly improving the emergency response capabilities and intelligence level of campus management.

[0016] As a further improvement to the technical solution of this invention, the preprocessed data is divided into three levels according to a preset business priority rule, specifically including: A pre-defined business priority rule base contains priority definitions for various types of data in different scenarios; Based on the data's scenario tags and data types, match the corresponding rules in the business priority rule library; Each data point is assigned a corresponding transmission level based on the matching results.

[0017] This invention pre-constructs a business priority rule base, which defines the transmission priority of various data types based on the business characteristics and data importance of different scenarios. After data preprocessing, the corresponding rules in the business priority rule base are automatically matched according to the scenario tags and data types carried by the data, assigning an appropriate transmission level to each data. This invention, through the pre-constructed business priority rule base, achieves automatic and accurate allocation of data transmission levels, ensuring the consistency and maintainability of the hierarchical transmission strategy. The rule base can be flexibly adjusted according to changes in campus business needs, adapting to changes in business priorities at different times and in different scenarios, further improving the flexibility and applicability of the hierarchical transmission mechanism.

[0018] As a further improvement to the technical solution of this invention, the specific steps of dynamically adjusting the cloud-edge task allocation strategy using reinforcement learning algorithms include: Use the global state information in the digital twin model as the input to the reinforcement learning agent; The optimization objectives are to minimize the overall system response time, maximize resource utilization, and minimize energy consumption. Through continuous learning and iteration of reinforcement learning agents, the optimal cloud-edge task allocation decision is output.

[0019] This invention uses the global state information of a digital twin model, including the resource status, network status, load status, and current list of tasks to be processed at each edge node, as the input state for the reinforcement learning agent. A reward function is defined with the shortest overall system response time, highest resource utilization, and lowest energy consumption as multi-objective optimization functions. Through continuous interaction and trial-and-error learning between the reinforcement learning agent and the environment, the task allocation strategy is continuously optimized, ultimately outputting the optimal cloud-edge task allocation decision. This invention employs a reinforcement learning algorithm for task scheduling, enabling it to autonomously learn system operating rules and dynamically optimize scheduling strategies without the need for manually pre-defined complex scheduling rules. The algorithm achieves a balance among multiple optimization objectives, not only improving overall system performance but also reducing system energy consumption, realizing intelligent and efficient scheduling of cloud-edge collaborative tasks.

[0020] As a further improvement to the technical solution of this invention, each edge node encrypts the trained model parameters and uploads them to the cloud server, specifically including: The model parameters obtained from edge node training are encrypted using a homomorphic encryption algorithm; The encrypted model parameters are uploaded to the cloud server through a secure communication channel; The cloud server aggregates model parameters in an encrypted state to generate global model parameters.

[0021] This invention employs a homomorphic encryption algorithm to encrypt the model parameters trained on edge nodes, ensuring that the model parameters remain encrypted throughout transmission. The encrypted model parameters are then uploaded to a cloud server via a secure communication channel. The cloud server, in the encrypted state, directly performs aggregation operations on all model parameters to generate encrypted global model parameters, which are then distributed to each edge node for decryption and updating. This invention achieves end-to-end encryption protection for model parameters during transmission and aggregation through homomorphic encryption technology. Even if the model parameters are intercepted during transmission, attackers cannot obtain any useful information. The cloud server completes parameter aggregation in an encrypted state without decrypting the original parameters, further enhancing the security of the federated learning process and effectively preventing attacks that use model parameters to deduce the original data.

[0022] Secondly, this invention provides a smart campus big data cloud-edge collaborative data acquisition and management system, comprising: The data acquisition module is used to collect multi-source heterogeneous data in various scenarios of smart campuses; The edge processing module is used to perform adaptive preprocessing and hierarchical transmission of the acquired multi-source heterogeneous data at the edge. The scheduling module is used to dynamically schedule cloud-edge collaborative tasks based on the digital twin virtual campus model. The analysis module is used to perform global intelligent analysis while protecting data privacy using a federated learning mechanism; The execution module is used to generate early warning signals for the operation status of the smart campus and drive the digital twin model to perform visualization.

[0023] This invention constructs a modular smart campus big data cloud-edge collaborative acquisition and management system, consisting of five core modules: acquisition, edge processing, scheduling, analysis, and execution. The acquisition module is responsible for the unified acquisition of multi-source heterogeneous data; the edge processing module is responsible for edge-side preprocessing and hierarchical transmission of data; the scheduling module is responsible for dynamic scheduling of cloud-edge tasks based on digital twins; the analysis module is responsible for global intelligent analysis based on federated learning; and the execution module is responsible for generating early warning signals and visualizing digital twins. The modules communicate with each other through standardized interfaces, collaboratively completing the entire data acquisition and management process. This invention adopts a modular design, with each module having independent functions and clear interfaces, facilitating system development, maintenance, and expansion. Modules can be flexibly added, removed, or upgraded according to actual needs, adapting to smart campus construction scenarios of different scales and requirements. The system as a whole achieves all the technical effects of the method described in this application, efficiently, securely, and intelligently completing the acquisition and management tasks of smart campus big data.

[0024] Compared with the prior art, the present invention has the following advantages: The smart campus big data cloud-edge collaborative acquisition and management method and system provided by this invention first collects multi-source heterogeneous data from multiple scenarios within the smart campus. This step enables unified access and standardized processing of data from various sensing devices within the campus, breaking down data silos between different scenarios and providing a comprehensive and accurate data foundation for subsequent data analysis. Second, the collected multi-source heterogeneous data undergoes edge-side adaptive preprocessing and hierarchical transmission. This step offloads some data processing tasks to edge nodes, reducing the amount of data that needs to be uploaded to the cloud, lowering network bandwidth pressure and the burden on cloud servers. Simultaneously, hierarchical transmission of data based on business priority ensures timely processing of business data with high real-time requirements and improves the utilization rate of network resources. Third, dynamic scheduling of cloud-edge collaborative tasks is achieved based on a digital twin virtual campus model. This step involves constructing a data twin that corresponds one-to-one with the physical campus. The digital twin model enables real-time perception and visualization of the overall campus status. Simultaneously, it employs reinforcement learning algorithms to dynamically adjust cloud-edge task allocation strategies, rationally distributing computational tasks based on edge node load and network conditions, thus improving system resource utilization and overall performance. Next, a federated learning mechanism is used to perform global intelligent analysis while protecting data privacy. This step achieves joint modeling of multiple edge nodes without disclosing raw data, solving the data silo problem and effectively protecting the privacy and security of students and faculty. Finally, a smart campus operation status early warning signal is generated and visualized using the digital twin model. This step can promptly detect abnormal states and potential risks during campus operation and provide intuitive visual alerts through the digital twin model, facilitating timely action by management personnel and improving the intelligence level and emergency response capabilities of campus management. In summary, the solution of this invention effectively solves the problems of high bandwidth pressure, high latency, high privacy and security risks, and low resource utilization in traditional centralized data processing models for smart campuses, achieving efficient, secure, and intelligent data collection and management for smart campuses. Attached Figure Description

[0025] Other features, objects, and advantages of the present invention will become more apparent from the following detailed description of non-limiting embodiments with reference to the accompanying drawings: Figure 1 This is an exemplary flowchart of a smart campus big data cloud-edge collaborative collection and management method according to some embodiments of the present invention; Figure 2 This is a schematic diagram illustrating an application scenario of the smart campus big data cloud-edge collaborative data collection and management system according to some embodiments of the present invention; Figure 3This is a schematic diagram of the edge-side adaptive preprocessing and hierarchical transmission process according to some embodiments of the present invention; Figure 4 This is a schematic diagram of the dynamic scheduling process for cloud-edge collaborative tasks according to some embodiments of the present invention; Figure 5 This is a schematic diagram of the structure of a smart campus big data cloud-edge collaborative data collection and management system according to some embodiments of the present invention; Figure 6 This is a schematic diagram of the structure of a computer device for implementing a smart campus big data cloud-edge collaborative collection and management method, as shown in some embodiments of the present invention. Detailed Implementation

[0026] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative effort are within the scope of protection of the present invention.

[0027] The present invention will be further described in detail below with reference to the accompanying drawings.

[0028] refer to Figure 1 The figure is an exemplary flowchart of a smart campus big data cloud-edge collaborative collection and management method according to some embodiments of the present invention. The smart campus big data cloud-edge collaborative collection and management method mainly includes the following steps: In step 101, multi-source heterogeneous data from multiple scenarios in the smart campus are collected.

[0029] In practical implementation, the collection of multi-source heterogeneous data across multiple scenarios in a smart campus can be achieved in the following ways: First, various sensing devices are deployed in scenarios such as teaching, security, energy consumption, logistics, and transportation within the smart campus. Specifically, sensing devices deployed in teaching scenarios include smart attendance machines, classroom environmental sensors, and multimedia equipment status sensors; sensing devices deployed in security scenarios include high-definition cameras, access control card readers, intrusion detection sensors, and fire alarms; sensing devices deployed in energy consumption scenarios include smart meters, smart water meters, smart gas meters, and air conditioning energy consumption sensors; sensing devices deployed in logistics scenarios include cafeteria payment machines, dormitory access control systems, and trash can overflow sensors; and sensing devices deployed in transportation scenarios include vehicle recognition cameras, parking space sensors, and shared bicycle locators. Then, structured, semi-structured, and unstructured data generated by these sensing devices are collected through a unified edge data access gateway. This edge data access gateway supports multiple industrial communication protocols, including but not limited to MQTT, CoAP, HTTP, Modbus, and BACnet, enabling seamless access for different types of sensing devices. Finally, a unique identifier, a timestamp, and a scene tag are added to each collected data. The unique identifier is used to distinguish different data records, the timestamp is used to record the precise time of data collection, and the scene tag is used to identify the campus scene to which the data belongs, thereby completing the collection of multi-source heterogeneous data in multiple scenarios of the smart campus. Other methods can also be used in other embodiments, which are not limited here.

[0030] In some embodiments, reference Figure 2 As shown in the figure, this is a schematic diagram of an application scenario of the smart campus big data cloud-edge collaborative data collection and management system according to some embodiments of the present invention. The figure includes three main components: the edge layer, the cloud layer, and the application layer. The edge layer includes various sensing devices and edge data access gateways, which are responsible for data collection and preliminary processing; the cloud layer includes cloud servers and a digital twin platform, which are responsible for global data storage, analysis, and task scheduling; the application layer includes a campus management platform, mobile terminals, and a digital twin visualization interface, which are responsible for providing users with various intelligent services.

[0031] In step 102, the collected multi-source heterogeneous data is subjected to edge-side adaptive preprocessing and hierarchical transmission.

[0032] In some embodiments, reference Figure 3 As shown in the figure, this is a schematic diagram of the edge-side adaptive preprocessing and hierarchical transmission process according to some embodiments of the present invention. In this embodiment, the edge-side adaptive preprocessing and hierarchical transmission of the collected multi-source heterogeneous data can be implemented by the following steps: In step 1021, the collected raw data is cleaned, denoised, format converted, and feature extracted.

[0033] In specific implementation, the cleaning, denoising, format conversion, and feature extraction of the collected raw data can be achieved in the following ways: First, the raw data is cleaned to remove duplicate, missing, and erroneous data; then, a filtering algorithm is used to denoise the data to remove noise interference; next, data of different formats are converted into a unified JSON format for easier subsequent processing and transmission; finally, key features are extracted from the preprocessed data, such as facial features from video data, peak energy consumption features from energy consumption data, and attendance time features from attendance data, thereby completing the edge-side adaptive preprocessing of the raw data. Other methods can also be used in other embodiments, which are not limited here.

[0034] In step 1022, the preprocessed data is divided into three levels according to the preset business priority rules: real-time critical data, important business data, and non-urgent statistical data.

[0035] In specific implementation, the preprocessed data can be divided into three levels according to preset business priority rules, which can be achieved in the following way: First, a preset business priority rule library is established, which contains priority definitions for various types of data in different scenarios. For example, fire alarm data and intrusion detection data in security scenarios are real-time critical data; attendance data in teaching scenarios and real-time energy consumption data in energy consumption scenarios are important business data; and garbage bin overflow statistics in logistics scenarios and monthly vehicle flow data in traffic scenarios are non-urgent statistics. Then, the corresponding rules in the business priority rule library are matched according to the data's scenario label and data type. Finally, a corresponding transmission level is assigned to each data based on the matching result. Other methods can also be used in other embodiments, and are not limited here.

[0036] In step 1023, real-time key data is directly uploaded to the cloud server, important business data is initially analyzed by the edge node and the analysis results are uploaded, and non-urgent statistical data is stored locally on the edge node and uploaded in batches periodically.

[0037] In specific implementation, the transmission methods for different levels of data can be implemented in the following ways: For real-time critical data, such as fire alarm data and intrusion detection data, due to their extremely high real-time requirements, edge nodes directly upload them to the cloud server to ensure that the cloud can obtain and process these data as soon as possible; For important business data, such as attendance data and real-time energy consumption data, edge nodes first perform preliminary analysis, such as counting class attendance and calculating real-time energy consumption growth rate, and then upload the analysis results to the cloud server. This reduces the amount of data transmission while ensuring that the cloud can obtain key business information; For non-urgent statistical data, such as garbage can overflow statistics and monthly vehicle traffic data, edge nodes store them in a local edge database and periodically upload them to the cloud server in batches during network idle periods (such as at night), thereby effectively utilizing network bandwidth resources and avoiding network congestion. Other methods can also be used in other embodiments, which are not limited here.

[0038] It should be noted that the above steps enable adaptive preprocessing and hierarchical transmission of data at the edge, delegating some data processing tasks to edge nodes, reducing the amount of data that needs to be uploaded to the cloud, reducing network bandwidth pressure and the burden on cloud servers. At the same time, hierarchical transmission of data according to business priority ensures that business data with high real-time requirements can be processed in a timely manner, while also improving the utilization rate of network resources.

[0039] In step 103, dynamic scheduling of cloud-edge collaborative tasks is realized based on the digital twin virtual campus model.

[0040] In some embodiments, reference Figure 4 As shown in the figure, this is a schematic diagram of the dynamic scheduling of cloud-edge collaborative tasks according to some embodiments of the present invention. In this embodiment, the dynamic scheduling of cloud-edge collaborative tasks based on the digital twin virtual campus model can be achieved by the following steps: In step 1031, a digital twin virtual campus model corresponding one-to-one with the physical campus is constructed to map the equipment status, environmental parameters and business operation status of the physical campus in real time.

[0041] In specific implementation, the construction of a digital twin virtual campus model corresponding one-to-one with the physical campus can be achieved in the following way: First, a three-dimensional geographic information model of the campus is constructed using 3D modeling technology, including the campus's buildings, roads, green spaces, and other infrastructure; then, various virtual entities are added to the 3D geographic information model, including sensing devices, teaching equipment, logistics equipment, personnel, etc., and these virtual entities correspond one-to-one with the entities in the physical campus; finally, the status information of various entities in the physical campus is synchronized to the corresponding virtual entities through a real-time data interface to achieve real-time mapping between the physical campus and the virtual campus. Other methods can also be used in other embodiments, which are not limited here.

[0042] In step 1032, the computing resources, storage resources, network bandwidth and load of each edge node are monitored in real time, and this status information is synchronized to the digital twin model.

[0043] In practice, real-time monitoring of the status information of each edge node can be achieved in the following way: deploy a status monitoring agent on each edge node, which collects status information such as CPU utilization, memory utilization, disk utilization, network bandwidth utilization, and task load of the edge node in real time; then, the status monitoring agent sends the collected status information to the cloud server periodically; finally, the cloud server synchronizes this status information to the digital twin virtual campus model, and displays the operating status of each edge node intuitively in the model. Other methods can also be used in other embodiments, which are not limited here.

[0044] In step 1033, based on the global state information in the digital twin model, a reinforcement learning algorithm is used to dynamically adjust the cloud-edge task allocation strategy, and the computing tasks are reasonably allocated to the cloud or edge nodes for execution.

[0045] In specific implementation, the reinforcement learning algorithm can be used to dynamically adjust the cloud-edge task allocation strategy in the following way: First, the global state information of the digital twin model, including the computing resources, storage resources, network bandwidth, load status of each edge node, and the current list of tasks to be processed, is used as the input to the reinforcement learning agent. Then, the reward function of reinforcement learning is defined with the optimization objectives of minimizing the overall system response time, maximizing resource utilization, and minimizing energy consumption. Next, the task allocation strategy is gradually optimized through continuous interaction and learning between the reinforcement learning agent and the environment. Finally, the reinforcement learning agent outputs the optimal cloud-edge task allocation decision, rationally allocating computing tasks to the cloud or edge nodes for execution. For example, when the load of an edge node is low and the network status is good, some computationally intensive tasks are assigned to that edge node for execution; when the load of an edge node is high or the network status is poor, tasks are assigned to the cloud server for execution. Other methods can also be used in other embodiments, and are not limited here.

[0046] It should be noted that the above steps can realize the dynamic scheduling of cloud-edge collaborative tasks based on the digital twin virtual campus model, which can reasonably allocate computing tasks according to global status information, thereby improving the system's resource utilization and overall performance.

[0047] In step 104, a federated learning mechanism is used to perform global intelligent analysis while protecting data privacy.

[0048] In practical implementation, the federated learning mechanism can be used to perform global intelligent analysis while protecting data privacy, as follows: First, machine learning models for specific scenarios are trained locally on each edge node. For example, edge nodes in a teaching scenario train student performance prediction models, edge nodes in a security scenario train abnormal behavior detection models, and edge nodes in an energy consumption scenario train energy consumption prediction models. The training process only uses data stored locally on the edge nodes and does not upload the raw data to the cloud. Next, each edge node encrypts the trained model parameters using a homomorphic encryption algorithm and uploads the encrypted model parameters to the cloud server through a secure communication channel. Then, the cloud server aggregates the model parameters uploaded by all edge nodes in the encrypted state to generate global model parameters. The aggregation process uses a federated averaging algorithm, i.e., a weighted average of the model parameters from all edge nodes. Finally, the cloud server distributes the updated global model parameters to each edge node, and each edge node uses the global model parameters to update its local model and continue the next round of training. Through multiple rounds of iterative training, a high-performance global model is finally obtained. This model integrates the data features of all edge nodes and can achieve more accurate global intelligent analysis. Other methods can also be used in other embodiments, which are not limited here.

[0049] It should be noted that the above steps adopt a federated learning mechanism, which enables joint modeling of multiple edge nodes without leaking the original data. This not only solves the data silo problem, but also effectively protects the privacy and security of students and faculty.

[0050] In step 105, a smart campus operation status early warning signal is generated and the digital twin model is used for visualization.

[0051] In practical implementation, generating early warning signals for the smart campus's operational status and driving the digital twin model for visualization can be achieved in the following way: First, based on global intelligent analysis results, the operational status of various scenarios within the smart campus is monitored in real time. For example, monitoring classroom attendance in teaching scenarios, personnel flow in security scenarios, energy consumption in energy consumption scenarios, and equipment operation in logistics scenarios. Then, when abnormal states or potential risks are detected, such as excessively low classroom attendance, abnormal gatherings of people on campus, a sudden and significant increase in energy consumption, or equipment malfunctions, an early warning signal is generated, including the location, type, severity, and suggested handling measures of the anomaly. Finally, the early warning signal is synchronized to the digital twin virtual campus model, driving the virtual entities in the corresponding areas of the model to display visual alarms such as highlighting and flashing, and displaying detailed early warning information in the virtual interface. For example, when a fire alarm is detected in a teaching building, the virtual entity of the teaching building in the digital twin model will turn red and flash continuously. At the same time, the specific location of the fire, the size of the fire, the evacuation route, and the suggested handling measures will be displayed in the virtual interface, so that the management personnel can take timely measures to deal with the situation. Other methods can also be used in other embodiments, which are not limited here.

[0052] It should be noted that the above steps can promptly identify abnormal states and potential risks in the campus operation process, and provide intuitive visual alerts through digital twin models, making it easier for managers to take timely measures to deal with them, thereby improving the level of intelligence in campus management and emergency response capabilities.

[0053] Furthermore, in another aspect, in some embodiments, the present invention provides a smart campus big data cloud-edge collaborative data acquisition and management system, as referenced. Figure 5 The figure is a schematic diagram of the structure of a smart campus big data cloud-edge collaborative acquisition and management system according to some embodiments of the present invention. The smart campus big data cloud-edge collaborative acquisition and management system includes: an acquisition module, an edge processing module, a scheduling module, an analysis module, and an execution module, which are described below: The data acquisition module in this invention is mainly used to collect multi-source heterogeneous data in various scenarios of smart campuses; The edge processing module in this invention is mainly used for adaptive preprocessing and hierarchical transmission of the collected multi-source heterogeneous data at the edge side. The scheduling module in this invention is mainly used to realize dynamic scheduling of cloud-edge collaborative tasks based on the digital twin virtual campus model; The analysis module in this invention is mainly used to perform global intelligent analysis while protecting data privacy using a federated learning mechanism. The execution module in this invention is mainly used to generate early warning signals for the operation status of the smart campus and drive the digital twin model to perform visualization.

[0054] The various modules in the aforementioned smart campus big data cloud-edge collaborative data collection and management system can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in the processor of a computer device in hardware form or independent of it, or stored in the memory of the computer device in software form, so that the processor can call and execute the corresponding operations of each module.

[0055] In another embodiment, the present invention provides a computer device, which may be a server, and its internal structure diagram may be as follows: Figure 6 As shown, the computer device includes a processor, memory, and a network interface connected via a system bus. The processor provides computing and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system, computer programs, and a database. The internal memory provides the environment for the operating system and computer programs in the non-volatile storage media to run. The database stores historical data, model parameters, and digital twin model data for smart campus big data cloud-edge collaborative acquisition and management. The network interface communicates with external terminals and edge nodes via a network connection. When the computer program is executed by the processor, it implements a smart campus big data cloud-edge collaborative acquisition and management method.

[0056] Those skilled in the art will understand that Figure 6 The structure shown is merely a block diagram of a portion of the structure related to the present invention and does not constitute a limitation on the computer device to which the present invention is applied. A specific computer device may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0057] In one embodiment, a computer device is also provided, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the steps in the above embodiment of the smart campus big data cloud-edge collaborative collection and management method.

[0058] In one embodiment, a computer-readable storage medium is provided, storing a computer program that, when executed by a processor, implements the steps described in the embodiment of the smart campus big data cloud-edge collaborative collection and management method.

[0059] In one embodiment, a computer program product or computer program is provided, comprising computer instructions stored in a computer-readable storage medium. A processor of a computer device reads the computer instructions from the computer-readable storage medium and executes the computer instructions, causing the computer device to perform the steps described in the embodiment of the smart campus big data cloud-edge collaborative acquisition and management method.

[0060] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium, and when executed, it can include the processes of the embodiments of the methods described above. Any references to memory, storage, databases, or other media used in the embodiments provided by this invention can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, or optical storage, etc. Volatile memory can include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM can be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM), etc.

[0061] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0062] The technical solutions provided by the embodiments disclosed in this invention have the following beneficial effects: The smart campus big data cloud-edge collaborative acquisition and management method and system provided by this invention first collects multi-source heterogeneous data from multiple scenarios within the smart campus. This step enables unified access and standardized processing of data from various sensing devices within the campus, breaking down data silos between different scenarios and providing a comprehensive and accurate data foundation for subsequent data analysis. Second, the collected multi-source heterogeneous data undergoes edge-side adaptive preprocessing and hierarchical transmission. This step offloads some data processing tasks to edge nodes, reducing the amount of data that needs to be uploaded to the cloud, lowering network bandwidth pressure and the burden on cloud servers. Simultaneously, hierarchical transmission of data based on business priority ensures timely processing of business data with high real-time requirements and improves the utilization rate of network resources. Third, dynamic scheduling of cloud-edge collaborative tasks is achieved based on a digital twin virtual campus model. This step involves constructing a data twin that corresponds one-to-one with the physical campus. The digital twin model enables real-time perception and visualization of the overall campus status. Simultaneously, it employs reinforcement learning algorithms to dynamically adjust cloud-edge task allocation strategies, rationally distributing computational tasks based on edge node load and network conditions, thus improving system resource utilization and overall performance. Next, a federated learning mechanism is used to perform global intelligent analysis while protecting data privacy. This step achieves joint modeling of multiple edge nodes without disclosing raw data, solving the data silo problem and effectively protecting the privacy and security of students and faculty. Finally, a smart campus operation status early warning signal is generated and visualized using the digital twin model. This step can promptly detect abnormal states and potential risks during campus operation and provide intuitive visual alerts through the digital twin model, facilitating timely action by management personnel and improving the intelligence level and emergency response capabilities of campus management. In summary, the solution of this invention effectively solves the problems of high bandwidth pressure, high latency, high privacy and security risks, and low resource utilization in traditional centralized data processing models for smart campuses, achieving efficient, secure, and intelligent data collection and management for smart campuses.

[0063] The embodiments described above are merely illustrative of several implementations of the present invention, and while the descriptions are relatively specific and detailed, they should not be construed as limiting the scope of the invention patent. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of the present invention, and these all fall within the protection scope of the present invention. Therefore, the protection scope of this invention patent should be determined by the appended claims.

Claims

1. A smart campus big data cloud-edge collaborative collection and management method, characterized in that, Includes the following steps: Collect multi-source heterogeneous data from various scenarios in smart campuses; Adaptive preprocessing and hierarchical transmission are performed on the edge side for the collected multi-source heterogeneous data; Dynamic scheduling of cloud-edge collaborative tasks based on a digital twin virtual campus model; A federated learning mechanism is used to perform global intelligent analysis while protecting data privacy. Generate early warning signals for the operation status of the smart campus and drive the digital twin model to visualize the display.

2. The smart campus big data cloud-edge collaborative collection and management method as described in claim 1, characterized in that, The collection of multi-source heterogeneous data in various scenarios of smart campuses specifically includes: Deploy various sensing devices in teaching, security, energy consumption, logistics, and transportation scenarios in smart campuses; The system collects structured, semi-structured, and unstructured data generated by various sensing devices through a unified edge data access gateway. Add a unique identifier, timestamp, and scene tag to each collected data.

3. The smart campus big data cloud-edge collaborative collection and management method as described in claim 1, characterized in that, The edge-side adaptive preprocessing and hierarchical transmission of the collected multi-source heterogeneous data specifically includes: The collected raw data is cleaned, denoised, format-converted, and feature-extracted. Based on preset business priority rules, the pre-processed data is divided into three levels: real-time critical data, important business data, and non-urgent statistical data. Real-time critical data is directly uploaded to the cloud server. Important business data is initially analyzed by edge nodes and the analysis results are uploaded. Non-urgent statistical data is stored locally on edge nodes and uploaded in batches periodically.

4. The smart campus big data cloud-edge collaborative collection and management method as described in claim 1, characterized in that, The dynamic scheduling of cloud-edge collaborative tasks based on the digital twin virtual campus model specifically includes: Construct a digital twin virtual campus model that corresponds one-to-one with the physical campus, and map the equipment status, environmental parameters and business operation status of the physical campus in real time; Real-time monitoring of computing resources, storage resources, network bandwidth and load of each edge node, and synchronization of this status information to the digital twin model; Based on the global state information in the digital twin model, a reinforcement learning algorithm is used to dynamically adjust the cloud-edge task allocation strategy, so as to reasonably allocate computing tasks to cloud or edge nodes for execution.

5. The smart campus big data cloud-edge collaborative collection and management method as described in claim 1, characterized in that, The use of federated learning mechanisms to perform global intelligent analysis while protecting data privacy specifically includes: Train machine learning models for specific scenarios locally on each edge node; Each edge node encrypts the trained model parameters and uploads them to the cloud server; The cloud server aggregates the model parameters uploaded by all edge nodes to generate a global model; The updated global model parameters are distributed to each edge node to guide the iterative optimization of the local model on the edge nodes.

6. The smart campus big data cloud-edge collaborative collection and management method as described in claim 1, characterized in that, Generating early warning signals for the operational status of the smart campus and driving the digital twin model to visualize them specifically includes: Based on the results of global intelligent analysis, the operational status of various scenarios in the smart campus is monitored in real time; When an abnormal state or potential risk is detected, an early warning signal is generated that includes the location, type, severity, and recommended handling measures of the abnormality. The warning signal is synchronized to the digital twin virtual campus model, which drives the virtual entities in the corresponding area of ​​the model to highlight and flash visual alarms, and displays detailed warning information in the virtual interface.

7. The smart campus big data cloud-edge collaborative collection and management method as described in claim 3, characterized in that, Based on preset business priority rules, the preprocessed data is divided into three levels, specifically including: A pre-defined business priority rule base contains priority definitions for various types of data in different scenarios; Based on the data's scenario tags and data types, match the corresponding rules in the business priority rule library; Each data point is assigned a corresponding transmission level based on the matching results.

8. The smart campus big data cloud-edge collaborative collection and management method as described in claim 4, characterized in that, The specific methods for dynamically adjusting cloud-edge task allocation strategies using reinforcement learning algorithms include: Use the global state information in the digital twin model as the input to the reinforcement learning agent; The optimization objectives are to minimize the overall system response time, maximize resource utilization, and minimize energy consumption. Through continuous learning and iteration of reinforcement learning agents, the optimal cloud-edge task allocation decision is output.

9. The smart campus big data cloud-edge collaborative collection and management method as described in claim 5, characterized in that, Each edge node encrypts the trained model parameters and uploads them to the cloud server, specifically including: The model parameters obtained from edge node training are encrypted using a homomorphic encryption algorithm; The encrypted model parameters are uploaded to the cloud server through a secure communication channel; The cloud server aggregates model parameters in an encrypted state to generate global model parameters.

10. A smart campus big data cloud-edge collaborative data acquisition and management system, characterized in that, include: The data acquisition module is used to collect multi-source heterogeneous data in various scenarios of smart campuses; The edge processing module is used to perform adaptive preprocessing and hierarchical transmission of the acquired multi-source heterogeneous data at the edge. The scheduling module is used to dynamically schedule cloud-edge collaborative tasks based on the digital twin virtual campus model. The analysis module is used to perform global intelligent analysis while protecting data privacy using a federated learning mechanism; The execution module is used to generate early warning signals for the operation status of the smart campus and drive the digital twin model to perform visualization.

Images