Smart campus resource scheduling method based on data fusion

Abstract

The invention relates to the technical field of smart campus resource scheduling, in particular to a smart campus resource scheduling method based on data fusion, which comprises the steps of setting a resource scheduling system with adjustable parameters, and introducing a multi-source heterogeneous data dynamic integration mechanism and a distributed resource allocation strategy. The influence of multi-user concurrent access on the resource allocation efficiency is quantified through a virtual user model, and a resource allocation efficiency model is constructed in combination with nonlinear regression analysis. The method can effectively solve the problems of insufficient resource scheduling flexibility, limited real-time performance and difficult consideration of security in a smart campus scene, provides theoretical support for efficient and intelligent resource configuration, and is suitable for multi-task cooperative processing and complex resource scheduling optimization in a high-load state.

Description

TECHNICAL FIELD

[0001] The present application belongs to the field of smart campus and data fusion technology, specifically a smart campus resource scheduling method based on data fusion. BACKGROUND

[0002] The continuous advancement of smart campus construction provides a broad application prospect for resource scheduling methods based on data fusion, which exhibits significant advantages in improving campus management efficiency and optimizing resource allocation. However, existing data fusion technologies and resource scheduling schemes still have certain limitations in processing multi-source heterogeneous data, achieving dynamic scheduling, and ensuring system security and real-time performance, making it difficult to fully meet the needs of complex scenarios in smart campus.

[0003] After searching, it was found that a data fusion system with publication number CN118764179B was disclosed on May 16, 2025. This patent uses a key server to perform obfuscation processing on data masks and perturbation perception data, and combines an operation server to generate a target obfuscated perception data sequence, finally completing data fusion. This technical solution has innovation in improving the security of perception data provided by data providers, but its focus is mainly on data security, and it lacks in the dynamicity and adaptability of resource scheduling. Specifically, this system does not fully consider the efficient integration and dynamic allocation strategy of multi-source heterogeneous data (such as teaching resources, equipment status, personnel activities, etc.) in the smart campus scenario, which may affect the flexibility and real-time performance of resource scheduling. In addition, this scheme has high demand for computing resources, which may increase the deployment and operation cost of the system.

[0004] Another patent document with publication number CN119296328B disclosed a traffic data fusion method and device on March 4, 2025. This technology realizes high-precision fusion of traffic data by performing grid processing on traffic condition data and combining morphological operations and connected component analysis, thereby improving traffic management efficiency. However, this technical solution is mainly aimed at the transportation field, and its data fusion method focuses on the construction of spatiotemporal traffic maps and the extraction of congestion coordinate sets, lacking comprehensive scheduling capabilities for diversified resources (such as classrooms, laboratories, network bandwidth, etc.) in the smart campus scenario. In addition, this scheme does not fully consider the situation of multi-user concurrent access and dynamic demand changes, which may limit its applicability in smart campus resource scheduling. At the same time, this method has high requirements for data preprocessing, which may increase the complexity and response delay of the system.

[0005] The aforementioned problems indicate that existing data fusion technologies, when applied to resource scheduling in smart campuses, still have room for improvement in areas such as efficient integration of multi-source heterogeneous data, formulation of dynamic scheduling strategies, and balancing system real-time performance and security. Therefore, this invention proposes a data fusion-based smart campus resource scheduling method, aiming to better meet the needs of efficient and intelligent resource scheduling in smart campus scenarios by optimizing data fusion algorithms, designing dynamic resource allocation mechanisms, and improving system real-time performance and security. Summary of the Invention

[0006] To address the aforementioned issues, this invention proposes a data fusion-based smart campus resource scheduling method. By introducing a dynamic integration mechanism for multi-source heterogeneous data and a distributed resource allocation strategy, a resource scheduling model suitable for complex scenarios is established. Furthermore, by quantifying the impact of concurrent access by multiple users on system response latency and combining high-precision experimental data analysis, the dynamic scheduling algorithm is designed and optimized. This solves the problems of insufficient resource scheduling flexibility, limited real-time performance, and difficulty in balancing security in smart campus scenarios, providing theoretical support for efficient and intelligent resource allocation in smart campus environments.

[0007] To achieve the above objectives, the technical solution of the present invention is implemented as follows:

[0008] A smart campus resource scheduling method based on data fusion includes:

[0009] A resource scheduling system with adjustable parameters is configured, and a multi-dimensional data acquisition module is introduced into the system.

[0010] The data acquisition module and resource allocation module are jointly configured to establish a basic data fusion link;

[0011] The collaboration between the data acquisition module and the resource allocation module is dynamically adjusted to fully activate the integration capability of multi-source heterogeneous data;

[0012] Virtual user models are deployed at intermediate nodes of the resource scheduling system to adjust the distance between virtual users and resource pools by setting a range, and multiple sets of resource allocation efficiency samples are obtained by segmented measurement.

[0013] By repeatedly measuring under multiple parameters, a resource allocation efficiency matrix containing two variables, virtual user distance and resource pool capacity, is constructed.

[0014] A resource allocation efficiency model with virtual user behavior compensation was established by using nonlinear regression analysis to perform bivariate joint fitting on the measurement data.

[0015] Further, the step specifically refers to: setting a data collection module DC, a resource allocation module RA, and a central processing unit CPU, wherein the cooperation parameters between the data collection module and the resource allocation module are adjustable, and the central processing unit is arranged between the data collection module DC and the resource allocation module RA.

[0016] Further, the step specifically refers to: the data collection module and the resource allocation module are connected to the central processing unit CPU through a high-speed communication interface, and the data collection module and the resource allocation module are both equipped with a data preprocessing unit for experiments, thereby establishing a basic data fusion link.

[0017] Further, the step specifically refers to: adjusting the cooperation parameters between the data collection module and the resource allocation module, thereby fully activating the integration capability of multi-source heterogeneous data.

[0018] Further, the step specifically refers to: fixing the communication bandwidth between the data collection module DC and the resource allocation module RA, adjusting the distance between the virtual user and the resource pool, collecting the resource allocation efficiency data of the virtual user at different distances by using the resource scheduling system, and calculating the average value of the resource allocation efficiency from the data as the final resource allocation efficiency value corresponding to each measurement.

[0019] Further, the step specifically refers to: changing the cooperation parameters between the data collection module and the resource allocation module and adjusting the value of the distance between the virtual user and the resource pool, repeating the above steps, and obtaining the demarcation value of the virtual user distance in the range of 0.1-1.0m; according to the multiple resource allocation efficiency values obtained in the above steps, a resource allocation efficiency model with the influence of the distance of the virtual user is constructed.

[0020] Further, the step specifically refers to: performing parameter fitting on the resource allocation efficiency model constructed in the above step by using the least square method to obtain the correction value of each parameter of the resource allocation efficiency model, and using the correction value of each parameter to obtain the final resource allocation efficiency model.

[0021] Further, in the step, the obtained resource allocation efficiency value is parameter fitted, and the resource allocation efficiency index can be expressed as:

[0022] Wherein, d represents the distance between the virtual user and the resource pool, and a, b, and c respectively represent the coefficients of each term of the above expression.

[0023] And the constructed resource allocation efficiency model with the influence of the distance of the virtual user is finally expressed as:

[0024] Wherein, d represents the distance between the virtual user and the resource pool; a, b, and c are the coefficients of each term of the above expression.

[0025] The expression of the final resource allocation efficiency model is represented as:

[0026] Wherein, is the selected reference distance, represents the resource allocation efficiency value at the reference distance, represents the cooperation parameter between the data acquisition module and the resource allocation module, represents the distance between the virtual user and the resource pool, represents the resource allocation efficiency index, and is the random noise obeying the normal distribution with the mean of 0 and the standard deviation of.

[0027] In order to achieve the above-mentioned purpose, the application further provides an application of a smart campus resource scheduling method based on data fusion. The smart campus resource scheduling method based on data fusion can be used in a multi-user concurrent access scene, or used as complex resource scheduling modeling of multi-task cooperative processing, or used as optimization of complex resource scheduling under high load state.

[0028] Beneficial effects: The application introduces a dynamic integration mechanism of multi-source heterogeneous data and a distributed resource allocation strategy, establishes an influence model of the distance between the virtual user and the resource pool on the resource allocation efficiency, quantizes the multi-dimensional changes caused by multi-user concurrent access, and combines high-precision experimental data analysis to complete the design and optimization of the dynamic scheduling algorithm, so as to solve the problems of insufficient resource scheduling flexibility, limited real-time performance and difficult-to-consider security in the smart campus scene, and provide theoretical support for efficient and intelligent resource configuration in the smart campus environment. In addition, the application can be used as complex resource scheduling modeling of multi-task cooperative processing (such as a hybrid scheduling mechanism combining teaching resource allocation and equipment state monitoring), and can be used as optimization of complex resource scheduling under high load state.

[0029] Figure 1 It is a schematic diagram of the overall architecture of the resource scheduling system in the embodiment of the application, which shows the connection relationship and cooperation process among the data acquisition module, the resource allocation module and the central processing unit.

[0030] Figure 2 It is a schematic diagram of the experimental scene of the virtual user model and the resource pool distance adjustment in the embodiment of the application, which shows the measurement process of the distance change between the virtual user and the resource pool on the resource allocation efficiency.

[0031] Figure 3 It is a flow chart of the construction of the resource allocation efficiency model in the embodiment of the application, which shows the process of fitting the final resource allocation efficiency model by repeated measurement and nonlinear regression analysis method under multi-parameter conditions.

[0032] The reference signs are as follows: 1, data acquisition module; 2, resource allocation module; 3, central processing unit; 4, virtual user model; 5, resource pool; 6, communication interface; 7, data preprocessing unit.

[0033] The present application provides a kind of wisdom campus resource scheduling method based on data fusion, its overall architecture is as shown in Figure 1 As shown in the figure. The system includes data acquisition module 1, resource allocation module 2, central processing unit 3, communication interface 6 and data preprocessing unit 7. Data acquisition module 1 and resource allocation module 2 are connected with central processing unit 3 through high-speed communication interface 6, form a complete resource scheduling link. Data acquisition module 1 is responsible for extracting information from multi-source heterogeneous data, which can include sensor network, user behavior log and equipment state monitoring etc. Resource allocation module 2 is used to dynamically allocate resource pool 5 according to the information provided by data acquisition module 1. Central processing unit 3 plays a coordinating role between the two, to ensure efficient transmission and processing of data flow. Data preprocessing unit 7 is respectively arranged in data acquisition module 1 and resource allocation module 2, for cleaning, formatting and preliminary analysis of raw data, to reduce the computational burden of subsequent processing.

[0034] In actual operation, first, the adjustable parameters of the system need to be initialized and configured. Specifically, the cooperation parameters between data acquisition module 1 and resource allocation module 2 are set to adjustable state, in order to optimize the data fusion capability in different scenarios. Cooperation parameters mainly include communication bandwidth, data sampling frequency and data transmission priority, etc. Figure 1 The figure clearly shows the connection relationship between these modules and their cooperation process. Central processing unit 3 as the core control unit, receives data from data acquisition module 1, and transmits it to resource allocation module 2 for further processing. In this process, communication interface 6 adopts high-speed serial bus design, to ensure the real-time and stability of data transmission. In addition, data preprocessing unit 7 runs independently inside each module, can effectively filter out noise data and correct outliers, so as to improve data quality.

[0035] In order to fully activate the integration capability of multi-source heterogeneous data, the system dynamically adjusts the cooperation relationship between data acquisition module 1 and resource allocation module 2. This process is realized by adjusting communication bandwidth and sampling frequency. For example, in high-load scenarios, the communication bandwidth is set to the maximum value, to ensure that data acquisition module 1 can quickly deliver a large amount of data to resource allocation module 2. At the same time, the sampling frequency will also be increased accordingly, to capture more detailed information. In low-load scenarios, the communication bandwidth and sampling frequency are appropriately reduced, to save system resources. This dynamic adjustment mechanism enables the system to maintain efficient operation under different load conditions. In addition, the cooperation parameters between data acquisition module 1 and resource allocation module 2 can also be adaptively optimized according to the historical data analysis results, to further improve the flexibility and adaptability of the system.

[0036] A virtual user model 4 is deployed in the middle node of the resource scheduling system to simulate the resource request behavior of real users. As shown in Figure 2 The distance between the virtual user model 4 and the resource pool 5 is set to an adjustable range of 0.1-1.0 meters. By adjusting the distance between the virtual user model 4 and the resource pool 5, the system can measure the resource allocation efficiency at different distances. During the experiment, the virtual user model 4 sends resource requests to the resource pool 5, and the resource allocation module 2 allocates resources according to the request content and records the allocation efficiency data. After each measurement is completed, the system will average a plurality of resource allocation efficiency data to obtain the final resource allocation efficiency value. This process is repeated at multiple distance values to obtain sufficient sample data for subsequent analysis.

[0037] Building a resource allocation efficiency matrix containing virtual user distance and resource pool capacity double variables is one of the key steps of the present application. Specifically, by changing the cooperation parameters between the data acquisition module 1 and the resource allocation module 2 and adjusting the distance between the virtual user model 4 and the resource pool 5, the system can obtain multiple sets of resource allocation efficiency data. These data are arranged in matrix form, where the rows represent the distance between the virtual user model 4 and the resource pool 5, and the columns represent the capacity of the resource pool 5. Each element in the matrix corresponds to the resource allocation efficiency value under the condition of a specific distance and capacity. Figure 3 The construction process of the resource allocation efficiency model is shown in detail, including data acquisition, matrix construction, and nonlinear regression analysis steps. Through nonlinear regression analysis method, the system can establish a resource allocation efficiency model with virtual user behavior compensation by jointly fitting the matrix data with two variables.

[0038] To further optimize the resource allocation efficiency model, the present application uses the least squares method to fit the measured data. Specifically, the resource allocation efficiency index can be expressed as a function expression, which contains the distance between the virtual user model 4 and the resource pool 5, the cooperation parameters between the data acquisition module 1 and the resource allocation module 2, and other related coefficients. By fitting multiple sets of experimental data, the system can obtain the corrected values of each coefficient, and use these corrected values to construct the final resource allocation efficiency model. The expression of the final model also includes a reference distance and a random noise term to more accurately reflect the actual resource allocation situation in the scene.

[0039] In practical applications, the resource scheduling method proposed by the present application can be used in multi-user concurrent access scenarios or multi-task cooperative processing scenarios. For example, in a hybrid scheduling mechanism that combines teaching resource allocation and device state monitoring, the system can dynamically adjust the resource allocation strategy according to user demand. When multiple users simultaneously request access to the same teaching resource, the system will prioritize allocation to the user with the most urgent demand based on the prediction results of the virtual user model 4. At the same time, the system also monitors the device state in real time to ensure that the devices in the resource pool 5 are always in a usable state. In high-load states, the system can effectively alleviate resource contention problems and improve overall operational efficiency by dynamically adjusting collaboration parameters and optimizing resource allocation efficiency models.

[0040] The present application not only applies to resource scheduling in a smart campus environment, but also can be extended to other complex scenarios. For example, during large conferences or events, the system can dynamically adjust resource allocation strategies based on the number of attendees and device usage to ensure that each user can have a high-quality service experience. In addition, the present application can also be applied to remote education platforms to improve the utilization of teaching resources and user experience by optimizing resource allocation strategies. By introducing a dynamic integration mechanism for multi-source heterogeneous data and a distributed resource allocation strategy, the present application successfully solves the problems of insufficient flexibility, limited real-time performance, and difficulty in balancing security in resource scheduling in a smart campus scenario, providing theoretical support and technical support for efficient and intelligent resource allocation. In order to better enable relevant personnel in the technical field to fully understand and implement the present application, the following further supplement the specific implementation principles of the present application in conjunction with a specific application scenario.

[0041] In a smart campus environment, dynamic allocation of resources such as classrooms, laboratories, and network bandwidth is a complex task. Suppose that a teaching building needs to simultaneously meet the teaching needs of multiple classes and allocate resources based on real-time device status and user behavior. First, the data acquisition module 1 extracts multi-source heterogeneous data from the sensor network, user behavior logs, and device state monitoring system. These data include but are not limited to the operating status of devices in the classroom, student learning behavior records, teacher teaching needs, etc. The data preprocessing unit 7 performs cleaning and formatting operations on the raw data to remove noise data and correct outliers, thereby providing high-quality basic data for subsequent data fusion and resource allocation.

[0042] Subsequently, the central processing unit 3 receives the processed data from the data acquisition module 1 and transmits it to the resource allocation module 2. In this process, the communication interface 6 is designed with a high-speed serial bus to ensure real-time and stability of data transmission. For example, in a high-load scenario, such as multiple classes simultaneously requesting the use of the same laboratory or network resources, the system dynamically adjusts the cooperation parameters between the data acquisition module 1 and the resource allocation module 2. Specifically, the communication bandwidth is set to the maximum value to support fast transmission of large amounts of data; at the same time, the sampling frequency is also increased accordingly in order to capture more detailed information. While in the low-load scenario, the communication bandwidth and sampling frequency are appropriately reduced to save system resources. This dynamic adjustment mechanism enables the system to maintain high efficiency under different load conditions.

[0043] To simulate the resource request behavior of real users, the system deploys a virtual user model 4 at the intermediate node. As shown in Figure 2 The distance between the virtual user model 4 and the resource pool 5 is set to an adjustable range, ranging from 0.1 to 1.0 meters. During the experiment, the virtual user model 4 sends resource requests to the resource pool 5, and the resource allocation module 2 allocates resources according to the request content and records the allocation efficiency data. After each measurement is completed, the system averages multiple sets of resource allocation efficiency data to obtain the final resource allocation efficiency value. This process is repeated at multiple distance values to obtain sufficient sample data for subsequent analysis. For example, when the distance between the virtual user model 4 and the resource pool 5 is 0.5 meters, the system records a higher resource allocation efficiency value, and when the distance increases to 1.0 meters, the resource allocation efficiency decreases. These data are arranged in matrix form, where the rows represent the distance between the virtual user model 4 and the resource pool 5, the columns represent the capacity of the resource pool 5, and each element in the matrix corresponds to the resource allocation efficiency value under specific distance and capacity conditions.

[0044] Next, a nonlinear regression analysis method is used to jointly fit the matrix data with two variables, establishing a resource allocation efficiency model with virtual user behavior compensation. Specifically, the resource allocation efficiency index can be expressed as a function expression, which includes the distance between the virtual user model 4 and the resource pool 5, the cooperation parameters between the data acquisition module 1 and the resource allocation module 2, and other related coefficients. By fitting multiple sets of experimental data, the system can obtain the corrected values of each coefficient and use these corrected values to build the final resource allocation efficiency model. The expression of the final model also includes a reference distance and a random noise term to more accurately reflect the actual resource allocation situation in the scene. For example, when the distance between the virtual user and the resource pool is 0.3 meters and the resource pool capacity is 80%, the system's resource allocation efficiency reaches the optimal value.

[0045] In practical applications, the resource scheduling method proposed by the present application can be used in multi-user concurrent access scenarios or multi-task cooperative processing scenarios. For example, in a hybrid scheduling mechanism combining teaching resource allocation and device state monitoring, the system can dynamically adjust the resource allocation strategy according to user demand. When multiple users simultaneously request access to the same teaching resource, the system will prioritize allocation to the user with the most urgent demand based on the prediction results of the virtual user model 4. At the same time, the system also monitors the device state in real time to ensure that the devices in the resource pool 5 are always in a usable state. In a high-load state, the system can effectively alleviate the resource contention problem by dynamically adjusting the cooperation parameters and optimizing the resource allocation efficiency model, thereby improving the overall operating efficiency.

[0046] In addition, the present application can also be applied to resource scheduling scenarios during large-scale conferences or events. For example, when an international conference is held in an academic report hall, the system can dynamically adjust the resource allocation strategy based on the number of participants and device usage to ensure that each user can have a high-quality service experience. For example, when the number of participants exceeds the predetermined threshold, the system will automatically increase the network bandwidth allocation and prioritize the normal operation of the main conference device. At the same time, the system will also adjust the state of environmental devices such as air conditioning and lighting based on real-time monitoring data to improve user experience.

[0047] In summary, by introducing a dynamic integration mechanism for multi-source heterogeneous data and a distributed resource allocation strategy, the present application successfully solves the problems of insufficient resource scheduling flexibility, limited real-time performance, and difficulty in balancing security in the smart campus scenario, providing theoretical support and technical support for efficient and intelligent resource allocation.

Claims

1. A smart campus resource scheduling method based on data fusion, characterized in that, Includes the following steps: S1: A resource scheduling system with adjustable parameters, incorporating a multi-dimensional data acquisition module; S2: Jointly configure the data acquisition module and the resource allocation module to establish a basic data fusion link; S3: Dynamically adjust the collaboration relationship between the data acquisition module and the resource allocation module to fully activate the integration capability of multi-source heterogeneous data; S4: Deploy a virtual user model at the intermediate node of the resource scheduling system to adjust the distance between the virtual user and the resource pool by setting a range, and obtain multiple sets of resource allocation efficiency samples using segmented measurement; specifically, step S4 refers to: fixing the communication bandwidth between the data acquisition module and the resource allocation module, adjusting the distance between the virtual user and the resource pool, using the resource scheduling system to collect resource allocation efficiency data of the virtual user at different distances, and calculating the average value of the resource allocation efficiency from this data as the final resource allocation efficiency value corresponding to each measurement; S5: Construct a resource allocation efficiency matrix containing two variables, virtual user distance and resource pool capacity, through repeated measurements under multi-parameter conditions; specifically, step S5 refers to: changing the cooperation parameters between the data acquisition module and the resource allocation module and adjusting the value of the distance between the virtual user and the resource pool, repeating the above steps, and obtaining the boundary value of the virtual user distance in the range of 0.1 to 1.0 meters; based on the above steps, obtain multiple resource allocation efficiency values ​​and construct a resource allocation efficiency model with the influence of virtual user distance; S6: Use nonlinear regression analysis to process the measurement data. A bivariate joint fitting is performed to establish a resource allocation efficiency model with virtual user behavior compensation. Specifically, step S6 involves fitting the parameters of the resource allocation efficiency model constructed in the previous steps using the least squares method to obtain the corrected values ​​of each parameter. The corrected values ​​are then used to obtain the final resource allocation efficiency model. In step S6, the obtained resource allocation efficiency values ​​are fitted with parameters, and the resource allocation efficiency index can be expressed as: where represents the distance between the virtual user and the resource pool, and , , and represent the coefficients of the above expression, respectively. The final expression of the constructed resource allocation efficiency model with the influence of virtual user distance is: where is the distance between the virtual user and the resource pool; are the coefficients of the above expression. The final expression of the resource allocation efficiency model is: where is the selected reference distance, represents the resource allocation efficiency value at the reference distance, represents the collaboration parameter between the data acquisition module and the resource allocation module, represents the distance between the virtual user and the resource pool, represents the resource allocation efficiency index, and is random noise following a normal distribution with a mean of 0 and a standard deviation of .

2. The smart campus resource scheduling method based on data fusion according to claim 1, characterized in that, The step S1 specifically refers to setting up a data acquisition module (1), a resource allocation module (2), and a central processing unit (3), wherein the cooperation parameters between the data acquisition module (1) and the resource allocation module (2) are adjustable, and the central processing unit (3) is placed between the data acquisition module (1) and the resource allocation module (2).

3. The smart campus resource scheduling method based on data fusion according to claim 1, characterized in that, The step S2 specifically refers to the following: the data acquisition module (1) and the resource allocation module (2) are connected to the central processing unit (3) through the high-speed communication interface (6), and both the data acquisition module (1) and the resource allocation module (2) are equipped with a data preprocessing unit (7) to conduct experiments, thereby establishing a basic data fusion link.

Images