A cloud computing-based teaching platform resource optimization system

Abstract

The application provides a cloud computing-based teaching platform resource optimization system, and belongs to the technical field of electronic digital data processing and product data management, and comprises the following steps: obtaining resource use data of a teaching platform through a resource collection module and transmitting the resource use data to a resource analysis module; obtaining feedback opinion data of students and teachers through a feedback module and processing the feedback opinion data to obtain a correlation matrix between each feedback data point and a resource state dimension; calculating a resource state index according to the resource use data and the correlation matrix through the resource analysis module, predicting the state of each resource dimension at a future time according to an analysis result; determining the priority of a resource demand end according to the analysis result through a resource scheduling module to generate a resource allocation priority table, judging inefficient resources according to the comparison between the use rate of the resource demand end and a preset threshold, identifying and releasing the inefficient resources, and adjusting the allocation of resources and bandwidth according to the resource allocation priority table and the released resources.

Description

Technical Field

[0001] This invention belongs to the field of electronic digital data processing and product data management technology, and in particular relates to a cloud computing-based teaching platform resource optimization system. Background Technology

[0002] With the rapid development of information technology, education is gradually shifting towards cloud computing platforms. Traditional education models struggle to meet the diverse and real-time learning needs of students. Due to the characteristics of teaching applications, the allocation of computing resources to meet teaching needs is a critical issue. However, current research does not consider the characteristics of application workloads, leading to unnecessary resource allocation and migration, and consequently, resource waste.

[0003] For example, the virtual machine management and scheduling method and system for campus cloud platforms disclosed in Chinese patent CN104239123B, while proposing a virtual machine scheduling method to address the periodicity, predictability, and batch nature of teaching applications on campus cloud platforms, effectively reducing the number of physical servers used and achieving load balancing in the physical machine cluster, neglects comprehensive resource optimization and scheduling. This results in the inability to meet personalized needs, restricts students' proactive learning ability, and is detrimental to student learning management.

[0004] In addition, the existing technology also has the following drawbacks:

[0005] 1. The utilization rate of course resources varies greatly at different times. Some course resources are prone to access delays during peak periods, while some low-demand course resources are left idle for a long time.

[0006] 2. The existing platform lacks an intelligent resource optimization mechanism, resulting in low resource utilization efficiency.

[0007] 3. Inappropriate resource allocation and slow response speed directly affect students' learning outcomes.

[0008] 4. Courses that have not been accessed for a long time still occupy the platform's storage, computing, and bandwidth resources, reducing resource utilization. Summary of the Invention

[0009] To address the widespread problems in this field, such as uneven resource allocation, complex resource management, lack of student learning management, inability to dynamically adapt and adjust resources, inability to allocate resources, and poor intelligence, this invention proposes a cloud computing-based teaching platform resource optimization system, which mainly includes:

[0010] Resource acquisition module, feedback module, resource analysis module, and resource scheduling module;

[0011] The resource acquisition module obtains resource usage data from the teaching platform and transmits it to the resource analysis module.

[0012] The feedback module includes an opinion collection unit and a feedback processing unit;

[0013] First, feedback data from trainees and instructors is obtained through the feedback collection unit. Then, the feedback processing unit processes the feedback data to obtain a correlation matrix between each feedback data point and the resource status dimension, and transmits it to the resource analysis module. The correlation matrix is ​​determined by calculating the covariance between the historical number of feedback opinions and the historical data of resource status. The covariance represents the strength of the association between feedback opinions and resource status.

[0014] The resource analysis module includes a resource analysis unit and a prediction unit.

[0015] The resource analysis unit calculates a resource status index based on the resource usage data and the correlation matrix. The resource status index is determined by a weighted combination of resource deviation and feedback deviation. The resource deviation represents the difference between the actual resource status and the ideal resource status, and the feedback deviation represents the impact of feedback on the resource status.

[0016] The prediction unit predicts the state of each resource dimension in the future based on the analysis results. The state in the future is determined by a linear combination of the current resource state and the rate of change of the resource state over time. If the state in the future is greater than a preset resource capacity threshold, it is marked as a high-demand resource and the resource scheduling module is triggered to allocate resources.

[0017] The resource scheduling module includes a resource allocation priority decision unit, a resource reclamation unit, and a dynamic scheduling unit;

[0018] The priority decision-making unit determines the priority of resource demand based on the analysis results and generates a resource allocation priority table. The priority is calculated by a weighted combination of task importance index and predicted excess demand. The task importance index is determined based on access frequency and activity level, and the predicted excess demand is determined by the difference between predicted resource demand and currently allocated resources.

[0019] The resource recycling unit detects the utilization rate of resources on the demand side, and identifies inefficient resources based on a comparison of the utilization rate with a preset threshold. If the utilization rate is less than the preset threshold, the resource is marked as an inefficient resource. The inefficient resource is then identified and released by the resource recycling unit.

[0020] The dynamic scheduling unit adjusts the allocation of resources and bandwidth according to the resource allocation priority table and the released resources, wherein the adjustment is determined based on the real-time resource status and the priority of high-demand tasks.

[0021] Optionally, in one embodiment of the present invention, the system further includes a change assessment module, which includes a change assessment unit and an early warning unit;

[0022] The resource status index (RSI) before and after a sampling period is obtained through the change assessment unit, and the resource change rate (ΔRSI) is calculated according to the following formula: In the formula, RSI(t-Δt) is the resource status index at time t-Δt, RSI(t) is the resource status index at time t, and Δt is the time interval;

[0023] The warning unit classifies the warning level based on the absolute value of the resource change rate ΔRSI. When the absolute value of the resource change rate is less than or below the change rate threshold 1, there is no warning; when the absolute value of the resource change rate is greater than or equal to the change rate threshold 1 and less than the change rate threshold 2, it is a low-level warning; when the absolute value of the resource change rate is greater than or equal to the change rate threshold 2, it is a high-level warning.

[0024] The beneficial effects achieved by this invention are:

[0025] 1. By combining the resource analysis module and the resource scheduling module, resources are dynamically allocated based on the analysis results, prioritizing high-demand tasks, ensuring task processing efficiency and system load balancing, and improving resource allocation efficiency.

[0026] 2. Through the cooperation of the resource scheduling module and the feedback module, user feedback is collected after resource scheduling, and the problems are fed back to the analysis module for strategy optimization to ensure that resource allocation is more in line with actual needs.

[0027] 3. By coordinating the feedback module and the resource acquisition module, user feedback data is transformed into optimized input for the acquisition module, improving the monitoring accuracy of resource usage data and ensuring that the system's prediction of future resource demand is more reliable.

[0028] 4. Through the cooperation of resource acquisition, resource analysis, resource scheduling and feedback modules, the system achieves full lifecycle management of resources, dynamically adjusts resource allocation, and optimizes resource utilization and task execution efficiency.

[0029] 5. By combining the resource acquisition module and the resource analysis module, the system can obtain resource usage information in real time and predict resource demand based on historical data, ensuring more accurate resource allocation and avoiding problems of over-allocation or insufficient resources. Attached Figure Description

[0030] The invention will be further understood from the following description taken in conjunction with the accompanying drawings. The components in the drawings are not necessarily drawn to scale, but rather the emphasis is on illustrating the principles of the embodiments. In different views, the same reference numerals designate the same parts.

[0031] Figure 1 This is a schematic diagram of the overall block shape of the present invention.

[0032] Figure 2 This is a block diagram of the resource scheduling module and teaching platform of the present invention.

[0033] Figure 3 This is a block diagram of the feedback module and resource analysis module of the present invention.

[0034] Figure 4 This is a schematic diagram of the workflow of the resource optimization system of the present invention. Detailed Implementation

[0035] The following specific embodiments illustrate the implementation of the present invention. Those skilled in the art can understand the advantages and effects of the present invention from the content disclosed in this specification. The present invention can be implemented or applied through other different specific embodiments, and various details in this specification can also be modified and changed based on different viewpoints and applications without departing from the spirit of the present invention. Furthermore, the accompanying drawings of the present invention are for simple illustrative purposes only and are not depictions of actual dimensions; this is stated beforehand. The following embodiments will further describe the relevant technical content of the present invention in detail, but the disclosed content is not intended to limit the scope of protection of the present invention.

[0036] Example 1: As Figure 1 , Figure 2 , Figure 3 and Figure 4 As shown in this embodiment, a cloud computing-based teaching platform resource optimization system mainly includes:

[0037] Resource acquisition module, feedback module, resource analysis module, and resource scheduling module;

[0038] The resource acquisition module obtains resource usage data from the teaching platform and transmits it to the resource analysis module.

[0039] The feedback module includes an opinion collection unit and a feedback processing unit;

[0040] First, feedback data from trainees and instructors is collected through the feedback collection unit. Then, the feedback processing unit processes the feedback data to obtain the correlation matrix between each feedback data point and the resource status dimension, and transmits it to the resource analysis module. The correlation matrix is ​​determined by calculating the covariance between the historical number of feedback opinions and the historical data of resource status. The covariance represents the strength of the association between feedback opinions and resource status.

[0041] The resource analysis module includes a resource analysis unit and a prediction unit;

[0042] The resource status index is calculated by the resource analysis unit based on resource usage data and correlation matrix. The resource status index is determined by a weighted combination of resource deviation and feedback deviation. Resource deviation represents the difference between the actual resource status and the ideal resource status, while feedback deviation represents the impact of feedback on the resource status.

[0043] The prediction unit predicts the state of each resource dimension in the future based on the analysis results. The state in the future is determined by a linear combination of the current resource state and the rate of change of the resource state over time. If the state in the future is greater than the preset resource capacity threshold, it is marked as a high-demand resource and the resource scheduling module is triggered to allocate resources.

[0044] The resource scheduling module includes a resource allocation priority decision unit, a resource reclamation unit, and a dynamic scheduling unit;

[0045] The priority decision-making unit determines the priority of resource demand based on the analysis results and generates a resource allocation priority table. The priority is calculated by a weighted combination of the task importance index and the predicted excess demand. The task importance index is determined based on the access frequency and activity level, and the predicted excess demand is determined by the difference between the predicted resource demand and the currently allocated resources.

[0046] The resource recycling unit detects the utilization rate of resources on the demand side and judges inefficient resources by comparing the utilization rate with a preset threshold. If the utilization rate is less than the preset threshold, it is marked as an inefficient resource. The resource recycling unit identifies and releases inefficient resources.

[0047] The dynamic scheduling unit adjusts the allocation of resources and bandwidth based on the resource allocation priority table and the released resources. The adjustment is determined according to the real-time resource status and the priority of high-demand tasks.

[0048] The resource acquisition module, resource analysis module, resource scheduling module, and feedback module are all deployed on the server, and the intermediate and control data of the resource acquisition module, resource analysis module, resource scheduling module, and feedback module are stored in the server's database.

[0049] The resource acquisition module collects resource usage data from the teaching platform; the feedback module collects feedback data from students and instructors and transmits it to the resource analysis module; the resource analysis module performs real-time analysis based on the collected resource usage data and the correlation matrix between each feedback data point and the resource status dimension obtained after processing by the feedback module to form analysis results and predict future resource needs; the resource scheduling module dynamically allocates resources according to the analysis results.

[0050] The teaching platform resource optimization system also includes a central processing unit (CPU). The CPU is connected to the resource acquisition module, resource analysis module, resource scheduling module, and feedback module. The CPU provides centralized control over these modules to improve the overall system's optimization and scheduling reliability.

[0051] In this embodiment, the correlation matrix M is:

[0052] In the formula, M i,j To assess the correlation strength between feedback question i and resource state dimension j, the value typically ranges from [0,1], M. i,j When M = 0, there is no correlation. i,j When = 1, then it is perfectly correlated.

[0053] The correlation strength is calculated according to the following formula:

[0054] In the formula, F i R is used to calculate the historical number of feedback questions. j For historical data on resource status, Cov(F) i R j Let be the covariance between the feedback problem and the resource status, and its value satisfies:

[0055] In the formula, n is the number of historical data points, and t is the time. For feedback question F i The mean, For resource state R j The mean, Here, it is assumed that the feedback data and resource usage data collected or gathered at the current time t include:

[0056] Feedback: "Loading time too long" (3 items), "Access failed" (2 items);

[0057] System monitoring: Bandwidth utilization: 85%, CPU utilization: 70%;

[0058] Then F1(t)=3, F2(t)=2, R1(t)=85%, R2(t)=70%.

[0059] Optionally, the resource analysis unit acquires the collected resource usage data and the correlation matrix between each feedback data point and the resource status dimension obtained after processing by the feedback module, and calculates the resource status index (RSI) according to the following formula:

[0060]

[0061] In the formula, D R (t) represents the resource deviation, D F (t) represents the feedback deviation.

[0062] Among them, resource deviation degree D R (t) is calculated according to the following formula:

[0063]

[0064] In the formula, R j (t) represents the actual resource status value, R j，opt The ideal resource status value is set by the system based on the monitored indicators and input from the human-computer interaction interface. n is the number of resource dimensions.

[0065] Among them, the feedback deviation D F (t) is calculated according to the following formula:

[0066]

[0067] In the formula, n is the number of resource dimensions, and W R，j (t) represents the comprehensive impact factor of the feedback question on the j-th resource dimension, where its value is determined according to the following formula:

[0068]

[0069] In the formula, m represents the total number of feedback issues reported by the system monitoring, and F i (t) represents the number of feedback questions for the i-th question, M i，j This is to assess the correlation strength between feedback question i and resource dimension j.

[0070] Optionally, the prediction unit obtains the analysis results from the resource analysis unit and predicts future resource demands according to the following formula, where the predicted state of the j-th resource dimension in the future t+Δt is...

[0071]

[0072] In the formula, R j (t) represents the current resource status. The rate of change of resource status over time;

[0073] If the predicted value If the resource capacity exceeds the system's set threshold range, it is marked as a high-demand resource, and the resource scheduling module is triggered to dynamically allocate computing resources and bandwidth.

[0074] By combining the resource acquisition module and the resource analysis module, the system can obtain real-time information on resource usage and predict resource demand based on historical data, ensuring more accurate resource allocation and avoiding problems such as over-allocation or resource shortage.

[0075] Optionally, the resource allocation priority decision unit obtains the analysis results, evaluates the resource allocation priority on the demand side, and generates a resource allocation priority table.

[0076] The priority of resource allocation is determined according to the following formula:

[0077]

[0078] In the formula, P j (t) is the resource allocation priority index for task or user j, U j (t) represents the importance index of task or user j, with a value range of [1, 10]. j (t) represents the predicted excess demand for task j, and RSI(t) represents the resource status index.

[0079] In this embodiment, the importance index U of task or user j j (t) Its value is determined by the following formula:

[0080]

[0081] In the formula, F j For the frequency of access to tasks or users (e.g., number of accesses per day), A j S represents the activity level of tasks or users (e.g., online time percentage). j The service level (e.g., regular, advanced user) for a task or user, whose value is set by the system.

[0082] The predicted excess demand D for task j j (t) is determined according to the following formula:

[0083]

[0084] In the formula, For the predicted resource requirements of task j, R j alloc (t) represents the amount of resources currently allocated to task j.

[0085] when This indicates that the task requires more resources; the excess is represented by D. j (t).

[0086] when For ≤R j alloc (t), the excess demand is 0.

[0087] Among them, the predicted resource demand of task j Calculate according to the following formula:

[0088]

[0089] In the formula, R j (t-1) represents the actual resource usage of task j at the previous time point, ΔR j (t) represents the predicted change in resource usage, satisfying:

[0090] ΔR j (t)=φ·(R j (t-1)-R j (t-2));

[0091] In the formula, φ is the adjustment coefficient for the trend of change. In this embodiment, an example of the value of the adjustment coefficient for the trend of change is provided, specifically:

[0092] 1) If the teaching platform is in normal operation (stable user access and no significant change in course resource demand), then the adjustment coefficient for the trend of change is φ = 0.2.

[0093] 2) During peak course registration periods (a large influx of users in a short period of time, resulting in a surge in course visits), the adjustment coefficient for the trend is φ = 0.8.

[0094] 3) In the event of a sudden event (such as a hacker attack) (a sudden surge in access leads to a sharp increase in resource demand in a short period of time), the adjustment coefficient for the trend is φ = 1.5.

[0095] Optionally, the resource recycling unit can detect the utilization rate of resources on the demand side, identify inefficient resources, and release inefficient resources to supply high-demand tasks.

[0096] Among them, the identification of inefficient resources is determined by the conditions for determining inefficient resources.

[0097] In this embodiment, the dynamic scheduling unit dynamically adjusts the resource allocation strategy by analyzing resource status and task requirements in real time, combined with resource allocation priority index, resource usage, and real-time feedback.

[0098] Specifically, resources are dynamically allocated based on the index from the resource allocation priority decision-making unit and the feedback results from the resource recycling unit. In situations of resource scarcity, priority is given to ensuring the resource needs of critical tasks or users.

[0099] Dynamic allocation avoids resource waste or improper allocation, thereby improving the overall performance of the system.

[0100] In this embodiment, a dynamic scheduling unit is provided to dynamically schedule resources, including the following steps:

[0101] S100, Real-time acquisition of resource and task status:

[0102] Collect the current usage status of resources, including: R k total (t) represents the total system resources, R k used (t) represents the amount of resources already allocated.

[0103] Additionally, it collects task or user status: U j (t) represents the importance index of task j, DR j (t) represents the resource deviation index of task j.

[0104] S101, Calculate the priority index P for task resource allocation. j (t):

[0105] The resource allocation priority index for each task is calculated using the following formula:

[0106]

[0107] In the formula, U j (t) represents the importance index of task j, DR j (t) is the resource deviation index of task j, satisfying:

[0108]

[0109] In the formula, R j,k (t) represents the actual usage of task j in resource dimension k, R j,k,opt Let n be the ideal usage of task j in resource dimension k, and n be the number of resource dimensions (such as CPU, memory, bandwidth, etc.).

[0110] Among them, the ideal usage R of task j in resource dimension k j,k,opt Determined according to the following formula:

[0111] R j,k,opt =R j,k,avg +α·σ j,k ;

[0112] In the formula, R j,k,avg Using the historical average, the following conditions must be met: In the formula, m is the number of historical data points (e.g., the past m time points), and R j,k(t) represents the actual usage of task j in resource dimension k at time t, and α is the adjustment coefficient, the value of which is determined by the system. Specifically, for periods of high resource demand fluctuation, α is increased; for periods of low resource demand fluctuation and stable usage, α is decreased; for sudden high loads, α is dynamically increased to quickly adapt to sudden resource demands; for off-peak periods, α is dynamically decreased to smooth resource allocation and reduce unnecessary frequent adjustments; σ j，k The historical usage volatility coefficient for task j on resource dimension k is calculated according to the following formula:

[0113]

[0114] In the formula, R j,k (t) represents the actual usage of task j in resource dimension k at time t, R j,k,avg Historical averages are used.

[0115] S102. Calculation of Total Resources and Surplus:

[0116] Calculate the remaining amount of resource dimension k:

[0117]

[0118] In the formula, R k available (t) represents the remaining available resources in resource dimension k, R k total (t) represents the total resource capacity of resource dimension k, and the actual resource capacity provided by the system, R. k used (t) represents the amount of resources used in resource dimension k, and its value is obtained in real time by monitoring the resource usage of tasks or users.

[0119] S103, Resource Allocation:

[0120] Resources are allocated according to a resource allocation priority index, that is: resources are allocated according to a resource allocation priority index, and the allocation of resources is determined by the following formula:

[0121]

[0122] In the formula, R j,k alloc (t+1) represents the amount of resources allocated to task j at time t+1, indicating the actual amount of resources allocated to task j in resource dimension k at time t+1, P j (t) is the resource allocation priority index for task j, R j,k base(t) represents the basic resource guarantee amount for task j, that is, the minimum guaranteed resource amount allocated by the system for task j in resource dimension k. In other words, regardless of how the priority of task resource allocation changes, the system will reserve basic resources for the task to ensure that critical tasks can operate normally. Its value is usually a fixed proportion or a fixed value. In this embodiment, the basic resource guarantee amount is allocated to the task according to a certain proportion of the total system resources. R j,k base (t)=ρ j,k *R k total (t), where ρ j,k R represents the basic guarantee ratio (e.g., 5%, 10%) for task j in resource dimension k. k total (t) represents the total amount of resources in resource dimension k;

[0123] For example: If the total bandwidth is R k total (t) = 1Gbps, and the basic guarantee ratio for a certain task is ρ. j,k =10%, then: R j,k base (t)=10%·1Gbps=100Mbps.

[0124] R k available (t) represents the remaining available resources in resource dimension k, whose value is used to monitor the total amount and usage of resources in real time, satisfying: R k available (t)=R k total (t)-R k used (t);

[0125] S104. Update the resource allocation table:

[0126] Update the resource allocation for task j:

[0127]

[0128] In the formula, R k used (t+1) represents the total amount of resources allocated or used in resource dimension k at time t+1, R k used (t) represents the total amount of resources allocated or used in resource dimension k at time t, R j,k alloc (t+1) represents the amount of resources newly allocated to task j in resource dimension k at time t+1.

[0129] By combining the resource analysis module and the resource scheduling module, resources are dynamically allocated based on the analysis results, prioritizing high-demand tasks, ensuring task processing efficiency and system load balancing, and improving resource allocation efficiency.

[0130] By working together with the resource scheduling module and the feedback module, user feedback is collected after resource scheduling, and issues are fed back to the analysis module for strategy optimization, ensuring that resource allocation is more in line with actual needs.

[0131] By combining the feedback module and the resource acquisition module, user feedback data is transformed into optimized input for the acquisition module, improving the monitoring accuracy of resource usage data and ensuring that the system's prediction of future resource demand is more reliable.

[0132] In addition, this invention also provides a workflow for a cloud-based teaching platform resource optimization system, including the following steps:

[0133] S1. Collect resource usage data from the teaching platform through the resource collection module;

[0134] S2. Collect feedback data from trainees and instructors through the feedback module and transmit it to the resource analysis module;

[0135] S3. The resource analysis module performs real-time analysis based on the collected resource usage data and the correlation matrix between each feedback data point and the resource status dimension obtained after processing by the feedback module to obtain the resource status index (RSI).

[0136] S4. The resource analysis module predicts the state of the j-th resource dimension in the future t+△t based on the Resource Status Index (RSI).

[0137] S5. If the predicted value is greater than the resource capacity threshold range set by the system, it is marked as a high-demand resource, and the resource scheduling module is triggered to dynamically allocate computing resources and bandwidth, and then jump to step S6; otherwise, repeat steps S1-S4.

[0138] S6. The resource scheduling module analyzes the resource allocation priority of the resource demand side based on the analysis results, identifies the resource utilization rate of the resource demand side and releases the resource, so as to adjust the resource allocation in real time.

[0139] Optionally, the workflow of the resource optimization system may also include:

[0140] In step S6, the resource scheduling module dynamically adjusts the resource allocation scheme according to the real-time status of high-demand resources to meet the needs of high-resource-allocation-priority tasks.

[0141] Optionally, the workflow of the resource optimization system further includes: in step S5, identifying inefficient resources is determined by the conditions for determining inefficient resources, and the conditions for determining inefficient resources are shown in the part of step S23 below.

[0142] Through the cooperation of the resource collection module, the resource analysis module, the resource scheduling module and the feedback module, the full life cycle management of resources is realized, the resource allocation is dynamically adjusted, and the resource utilization rate and task execution efficiency are optimized.

[0143] Embodiment 2: This embodiment should be understood as including all the features of any one of the foregoing embodiments, and further improved on this basis. According to Figure 1 、 Figure 2 、 Figure 3 、 Figure 4 shown, it also includes a change evaluation module. The change evaluation module evaluates the change rate of the demand side according to the analysis results in a sampling period, triggers a warning prompt to the manager according to the change of the change rate, and dynamically adjusts the resource scheduling;

[0144] The change evaluation module includes a change evaluation unit, a warning unit, and an optimization unit. The change evaluation unit obtains the resource status index RSI before and after a sampling period, and calculates the resource change rate △RSI according to the following formula:

[0145]

[0146] In the formula, RSI(t - Δt) is the resource status index at time t - Δt, RSI(t) is the resource status index at time t, and Δt is the time interval;

[0147] [[ID=2,6]]The warning unit classifies the warning levels according to the following formula:

[0148] That is, classify according to the absolute value of the resource change rate:

[0149] If the absolute value of the resource change rate is less than the change rate threshold 1, that is, ∣△RSI∣ < Threshold1, it is no warning.

[0150] If the absolute value of the resource change rate is greater than or equal to the change rate threshold 1 and less than the change rate threshold 2, that is, Threshold1 ≤ ∣△RSI∣ < Threshold2, it is a low-level warning.

[0151] If the absolute value of the resource change rate is greater than or equal to the change rate threshold 2, that is, ∣△RSI∣ ≥ Threshold2, it is a high-level warning.

[0152] The warning unit triggers a warning prompt according to the warning classification. Specifically:

[0153] Low-level alert: Send learning status reminders to the demand side (e.g., "Learning status fluctuates greatly, please check your learning plan").

[0154] Advanced alerts: Notify demanders and managers to automatically recommend targeted resources (such as supplementary review materials, emotional management, and other mental health courses).

[0155] The change assessment module enables managers to grasp the trend of demand-side changes in real time, plan intervention measures in advance, and ensure that the entire system has the flexibility to dynamically adapt to demand-side needs.

[0156] Meanwhile, the specific values ​​of Threshold1 and Threshold2 need to be set according to the actual use scenario and input from the human-computer interaction interface, which will not be elaborated here.

[0157] When the demand side is in an advanced warning state, the optimization unit is triggered to dynamically adjust resource scheduling to ensure intelligent resource recovery and release.

[0158] The optimization unit reclaims and releases resources according to the following steps:

[0159] S21. Real-time collection of resource usage data;

[0160] The data collected on resource usage includes:

[0161] Resource usage R of task or user j at time t j (t), the amount of resources R currently allocated to task or user j. j alloc (t); Calculate the resource utilization rate RE according to the following formula. j (t):

[0162] S22. Assess resource status:

[0163] Calculate the resource deviation index DR for task j j (t), and calculate the average utilization rate The average utilization rate is calculated according to the following formula:

[0164] S23, Trigger inefficient resource flag:

[0165] Check if the inefficiency criteria are met:

[0166] 1)RE j (t) <Threshold Utilization ;2)DR j (t)>Threshold Deviation ;3)

[0167] Resources that meet the criteria are marked as inefficient.

[0168] Among them, Threshold Deviation and Threshold Utilization The system configures these settings based on actual conditions and inputs them through the human-computer interaction interface. Specifically, Threshold... Utilization The threshold is set by the system and its value range is [0.1, 0.7]. Specifically, in this embodiment, specific value examples are given: 1) If the teaching platform is operating normally (user access and use of course resources are relatively stable, and some resources are allowed to run inefficiently for a short period of time), then Threshold is set to [0.1, 0.7]. Utilization =0.3; 2) During peak course periods (a large number of users accessing the site simultaneously in a short period of time, resulting in resource scarcity), Threshold... Utilization =0.5; 3) Sudden events (the system experiences abnormally high loads, and resource supply and demand are tight), then Threshold Utilization =0.7.

[0169] S24. Release inefficient resources:

[0170] If the number of marked resources is small, they are released and directly allocated to tasks with higher resource allocation priority. If the number of marked resources is large, they are released gradually according to resource allocation priority. The amount of resources released is R. j release (t) is calculated according to the following formula:

[0171]

[0172] In the formula, R j (t) represents the resource usage of task or user j at time t, R j alloc (t) represents the amount of resources currently allocated to task or user j;

[0173] S25. Reallocate resources:

[0174] Based on the high resource allocation priority task P j (t) Allocate and release resources;

[0175] S26. Update resource status: Update the system's resource allocation table.

[0176] Through dynamic analysis and evaluation by the change assessment unit, combined with the intelligent triggering mechanism of the early warning unit and the efficient resource scheduling of the optimization unit, resource utilization is maximized, resource waste is effectively reduced, and the entire system is guaranteed to have the advantages of intelligent management, high flexibility and responsiveness, improved resource utilization, optimized user experience, and reduced operating costs.

[0177] The content disclosed above is merely a preferred embodiment of the present invention, and the description of the above embodiments is only for the purpose of helping to understand the method and core idea of ​​the present invention; at the same time, those skilled in the art will recognize that, based on the idea of ​​the present invention, there will be changes in specific implementation methods and application scope. In summary, the content of this specification should not be construed as a limitation of the present invention.

Claims

1. A cloud computing-based teaching platform resource optimization system, characterized in that, Comprise: Resource acquisition module, feedback module, resource analysis module and resource scheduling module; Through the resource acquisition module to obtain the resource use data of the teaching platform, and transmit to the resource analysis module; The feedback module includes opinion collection unit and feedback processing unit; Firstly, the feedback opinion data of the students and teachers is obtained through the opinion collection unit, and then the correlation matrix between each feedback data point and resource state dimension is obtained by processing the feedback opinion data through the feedback processing unit, and is transmitted to the resource analysis module; The correlation matrix is determined by the covariance calculation of the feedback opinion history quantity and the resource state history data, and the covariance represents the correlation strength of the feedback opinion and the resource state; Through the resource analysis module including resource analysis unit and prediction unit; Through the resource analysis unit, the resource state index is calculated according to the resource use data and the correlation matrix, wherein the resource state index is determined by the weighted combination of resource deviation and feedback deviation, the resource deviation represents the difference between actual resource state and ideal resource state, and the feedback deviation represents the influence of feedback opinion on resource state; Through the prediction unit, the state of each resource dimension in the future time is predicted according to the analysis result, wherein the state of the future time is determined by the linear combination of the current resource state and the resource state time change rate; If the state of the future time is greater than the preset resource capacity threshold, it is marked as high demand resource and the resource scheduling module is triggered to allocate resources; The resource scheduling module includes resource allocation priority decision unit, resource recycling unit and dynamic scheduling unit; Through the priority decision unit, the priority of the resource demand end is determined according to the analysis result to generate a resource allocation priority table, wherein the priority is calculated by the weighted combination of task importance index and predicted over demand amount, the task importance index is determined according to the access frequency and activity, and the predicted over demand amount is determined by the difference between the predicted resource demand and the current allocated resource; Through the resource recycling unit, the usage rate of the resource demand end is detected, and the inefficient resource is judged according to the comparison between the usage rate and the preset threshold, wherein if the usage rate is less than the preset threshold, it is marked as inefficient resource; The inefficient resource is identified and released by the resource recycling unit; Through the dynamic scheduling unit, the allocation of resources and bandwidth is adjusted according to the resource allocation priority table and the released resources, wherein the adjustment is determined according to the real-time resource state and the priority of high demand task.

2. The teaching platform resource optimization system of claim 1, wherein, The resource state index is denoted as RSI, and the calculation formula is: In the formula, D R (t) is the resource deviation degree, D F (t) is the feedback deviation degree; Wherein, the calculation formula of resource deviation degree is: In the formula, R j (t) is the actual resource state value, R j,opt is the ideal resource state value, which is set by the system according to the monitored indicators and input from the human-computer interaction interface, and n is the number of resource dimensions. Feedback deviation D F The calculation formula of (t) is: In the formula, n is the number of resource dimensions, W R,j (t) is the comprehensive influence factor of the feedback problem on the jth resource dimension. W R,j The calculation formula of (t) is: In the formula, m is the total number of feedback problems monitored by the system, F i (t) is the number of the i-th feedback problem, M i，j is the correlation strength of feedback problem i and resource dimension j.

3. The teaching platform resource optimization system of claim 1, wherein, The predicted value of the state at the future time is denoted as The calculation formula is: In the formula, R j (t) is the current resource state, is the time variation rate of the resource state.

4. The teaching platform resource optimization system of claim 1, wherein, The correlation matrix is denoted by M, which is expressed as: In the formula, M i,j is the correlation strength of the feedback question i and the resource state dimension j, and is usually in the range of [0, 1], M i,j = 0, then there is no correlation, M i,j = 1, then it is completely correlated; Correlation strength M i,j The calculation formula is: In the formula, F i is the historical number of feedback problems, R j is the historical data of the resource state, Cov(F i , R j ) is the covariance of the feedback problem and the resource state, and the value satisfies: In the formula, n is the number of time points of the historical data, t is time, is the mean of the feedback problem F i , is the mean of the resource state R j , 5. The teaching platform resource optimization system of claim 1, wherein, The priority of the resource demand end is calculated according to the following formula: In the formula, P j (t) is the resource allocation priority index of the task or user j, U j (t) is the importance index of the task or user j, and the value range is [1, 10], D j (t) is the predicted over-demand amount of the task j, and RS I(t) is the resource state index. importance index U of a task or user j j The calculation formula of U(t) is as follows: In the formula, F j is the access frequency of a task or user, A j is the activity of a task or user j, S j is the service level of a task or user j, whose value is set by the system; and D j The calculation formula of U(t) is as follows: wherein, Rj(t) is the predicted resource requirement of task j, j alloc (t) is the amount of resources currently allocated to task j.

6. The teaching platform resource optimization system of claim 1, wherein, The system further comprises a change evaluation module, which comprises a change evaluation unit and a warning unit; The resource state index RSI before and after the sampling period is obtained by the change evaluation unit, and the resource change rate ΔRSI is calculated according to the following formula: In the formula, RSI(t-Δt) is the resource state index at time t-Δt, RSI(t) is the resource state index at time t, and Δt is the time interval. Through the warning unit, the absolute value of resource change rate△RSI is classified according to the warning level, when the absolute value of resource change rate is less than the change rate threshold 1, there is no warning; The absolute value of resource change rate is greater than or equal to the change rate threshold 1 and less than the change rate threshold 2, which is low level warning; The absolute value of resource change rate is greater than or equal to the change rate threshold 2, which is high warning.

Images