Multi-campus linkage teaching resource blockchain dynamic scheduling system

Abstract

The application provides a multi-campus linkage teaching resource blockchain dynamic scheduling system, relates to the technical field of blockchain, and solves the problem of inaccurate resource demand estimation; a classroom type preference matrix is constructed through principal component analysis, type demand coefficients are generated in combination with expert experience and user feedback, accurate matching of resource allocation and actual demand is realized; through the collection and processing of real-time data points in the edge node, the multi-chain architecture maintains the consistency of the metadata of each classroom and the intelligent event triggering, reduces the influence of scheduling misjudgment and frequent compensation on stability; a federal learning dynamic optimization resource prediction model is adopted to protect data privacy and improve performance, the data optimization and adjustment accuracy are improved through energy consumption smoothing factor updating; the blockchain governance mechanism ensures the encryption of optimized data and the real-time synchronization of edge modules, reduces the misjudgment rate, improves resource management and equipment stability, and optimizes user experience.

Description

Technical Field

[0001] This invention relates to the field of blockchain technology, specifically a multi-campus collaborative teaching resource blockchain dynamic scheduling system. Background Technology

[0002] In the prior art, publication number CN116582704B, entitled "A Virtual Simulation Teaching Resource Sharing Management Platform," discloses a virtual simulation teaching resource sharing management platform. This platform, belonging to the field of virtual simulation teaching technology, addresses the problem of resource disparities in current virtual simulation training labs. It includes a comprehensive monitoring module, a scheduling analysis module, a training management system, a resource scheduling module, and a training feedback module. The scheduling analysis module analyzes the teaching resource scheduling needs of the virtual simulation training lab to obtain resource scheduling. The comprehensive monitoring module monitors the usage of the virtual simulation training lab. The resource scheduling module analyzes the scheduling of computing resources in each virtual simulation training lab to obtain the resource scheduling compensation level. The training feedback module analyzes the usage feedback of the virtual simulation training lab to obtain the training feedback level. This invention achieves rational allocation and intelligent sharing of virtual simulation teaching resources based on teaching resource needs.

[0003] However, in practical applications, when virtual simulation teaching resource sharing and management platforms are used in classrooms across multiple campuses, such as piano rooms, laboratories, and library study rooms, the following technical drawbacks often exist:

[0004] 1. The training configuration data lacks detail for different classroom types, resulting in inaccurate resource demand estimates.

[0005] 2. Misjudgments in resource scheduling lead to frequent compensation, resulting in overload of computing resources and frequent switching of power lines, affecting equipment stability and user experience.

[0006] The above-mentioned technical shortcomings create a chain reaction, affecting the resource management and utilization of classrooms across multiple campuses;

[0007] The information disclosed in the background section above is only intended to enhance the understanding of the background of this disclosure, and therefore may include information that does not constitute prior art known to those skilled in the art. Summary of the Invention

[0008] The purpose of this invention is to provide a blockchain-based dynamic scheduling system for teaching resources across multiple campuses, in order to solve the problems mentioned in the background section.

[0009] To achieve the above objectives, the present invention provides the following technical solution:

[0010] A multi-campus collaborative teaching resource blockchain dynamic scheduling system is applied to the sharing of high-demand resources among adjacent campuses, including:

[0011] The classroom type identification module is used to create a classroom type preference matrix and, during system initialization and periodic updates, to identify the usage characteristics of classrooms in each campus through the classroom type preference matrix and calculate the corresponding type demand coefficient Dxq.

[0012] The data acquisition and processing module collects real-time environmental status data, computing power data, and power demand data of classrooms at the edge nodes of each campus; it also classifies, segments, and homomorphically encrypts the raw data for different classroom types, and generates data blocks with type labels.

[0013] The blockchain storage consensus module is used to store data blocks marked with each classroom type; for each write operation, an associated smart event is generated and a threshold value M is set to trigger a "data ready" signal in real time; at the same time, a computing power demand fluctuation model is built to analyze the historical fluctuation range of each type of classroom.

[0014] The intelligent scheduling and execution module, upon receiving "data ready," reads the resource configuration parameters of each classroom and the latest fluctuations in computing power demand from the blockchain; simultaneously, it calculates the computing power priority allocation value P and generates a smoothed resource allocation scheme.

[0015] Write the resource allocation plan into the blockchain to trigger the execution of computing power and electricity, and monitor the switching of computing power resources and electricity in real time.

[0016] The feedback optimization and federated learning module sets an optimization cycle and uses federated learning to train the "resource prediction and certification performance model" in parallel on each campus, and calculates the updated energy consumption smoothing factor S. It is then distributed to the edge acquisition and scheduling execution module through on-chain governance to further improve prediction accuracy, reduce the frequency of misjudgments, and reduce the number of power supply switching times.

[0017] Furthermore, the classroom type recognition module specifically includes:

[0018] By collecting historical data on classroom usage frequency, usage duration, and user satisfaction in each classroom, principal component analysis is used to extract features from the historical data, including acoustic requirements, equipment requirements, and spatial requirements.

[0019] Based on the extracted features, a classroom type preference matrix is ​​constructed, with different types of classrooms as rows and each feature requirement as a column. Specific feature weights are generated to form the classroom type preference matrix.

[0020] The classroom types include piano rooms, laboratories, and library study rooms; the classroom type preference matrix consists of the acoustic requirements, connectivity requirements, and environmental stability of each type of classroom.

[0021] After identifying the characteristics of classrooms in each campus, the feature weights are multiplied by the actual performance values ​​of the corresponding classrooms, and then the results of all features are summed up to obtain the final type demand coefficient Dxq.

[0022] Furthermore, the data acquisition and processing module includes a data acquisition unit:

[0023] The data acquisition unit collects real-time environmental status data, computing power data, and power demand data of classrooms at the edge nodes of each campus.

[0024] The environmental data includes classroom temperature, humidity, air quality, and lighting intensity;

[0025] Computing power data includes processor utilization, memory usage, and network bandwidth of computer equipment;

[0026] Electricity demand data includes classroom electricity consumption and peak load;

[0027] The collected environmental status data, computing power data, and power demand data are transmitted using an encrypted transmission protocol.

[0028] Furthermore, the data acquisition and processing module also includes a threshold stratification processing unit;

[0029] The threshold layering processing unit is used to receive environmental status data, computing power data, and power demand data, and process them through a funnel filter window to filter out noise data.

[0030] Based on different classroom types, threshold criteria for stratification are set, including a first threshold and a second threshold, to stratify the data. Specifically:

[0031] For different classroom types, different stratification criteria thresholds are preset at the campus-level edge nodes; baseline thresholds are calculated based on historical data, including the arithmetic mean and standard deviation of the data during the statistical period.

[0032] The thresholds for each classroom type are generated and updated in real time. The threshold stratification processing unit continuously calculates the new arithmetic mean and standard deviation and applies an exponentially weighted moving average to smooth the thresholds.

[0033] The real-time data points after funnel filtering are classified into layers. By comparing and evaluating the real-time data points with the layer classification criteria threshold, they are labeled separately, including the first load layer, the second load layer, and the third load layer, thereby generating data blocks with hierarchical labels.

[0034] Furthermore, the blockchain storage consensus module includes multi-chain architecture units:

[0035] The multi-chain architecture unit is used to set up campus-level edge nodes; an independent sidechain is configured for each classroom type, and a classroom type tag is written in the genesis block of each sidechain;

[0036] For each write operation of the classroom type label, an associated smart event is automatically generated based on the characteristics of the written data; the smart event contains the data's metadata information, namely metadata type, timestamp, data block identifier, and cumulative signature.

[0037] A threshold value M is preset to confirm the number of nodes. When the number of confirmed nodes reaches or exceeds the threshold value M, a "data ready" signal is triggered, and the written data and its associated smart events are encrypted and stored in the corresponding sidechain.

[0038] Furthermore, the blockchain storage consensus module also includes a demand fluctuation model construction unit;

[0039] The demand fluctuation model construction unit collects historical data on resource usage, including the frequency and time of data writing and access for each type of classroom, and inputs this data into the computing power demand fluctuation model.

[0040] Time series analysis is used to predict future data access and write requirements for each type of classroom; the time series model takes into account historical periodic and trend changes to generate a forecast report for each classroom type.

[0041] Based on the prediction results, the resource allocation of each node in the blockchain network is dynamically adjusted; specifically, the support of nodes is automatically increased for the types of classrooms expected during peak demand periods, while the resource allocation is reduced during off-peak demand periods.

[0042] In addition, the demand fluctuation model will monitor in real time the difference between actual data usage and forecasts, and automatically adjust the historical periodic and trend changes for the next monitoring period.

[0043] Furthermore, the intelligent scheduling execution module includes a resource scheduling unit;

[0044] After receiving the "data ready" signal transmitted in the blockchain network, the resource scheduling unit instantly accesses the resource configuration parameters of each type of classroom and its latest computing power demand fluctuation information from the blockchain database.

[0045] Based on the obtained resource configuration parameters and the latest computing power demand fluctuation information, the resource demand priority of each type of classroom is calculated using a computing power priority allocation algorithm;

[0046] Generate a smoothed resource allocation scheme based on the calculated priority allocation value P to adjust the resource allocation for different types of classrooms.

[0047] Furthermore, the intelligent scheduling and execution module also includes a resource management unit;

[0048] The resource management unit records the resource allocation scheme reconfigured according to the priority allocation value P to the blockchain network, triggering the automatic execution of computing power and power resources;

[0049] During the process, it is controlled by a pre-defined smart contract, which simultaneously monitors and adjusts the switching between computing resources and electricity in real time.

[0050] Furthermore, the feedback optimization and federated learning module specifically includes:

[0051] The configuration and launch of federated learning are used to jointly optimize and train resource prediction and certification performance models without sharing campus data, thereby protecting data privacy while improving the global performance of the model.

[0052] The energy consumption smoothing factor S is calculated and updated using the federated learning results, employing the following update formula:

[0053]

[0054] in, α represents the smoothing parameter, where 0 < α < 1, Sold is the previous energy consumption smoothing factor, and Slear is the updated value derived from the current federated learning results.

[0055] Furthermore, the feedback optimization and federated learning module also includes:

[0056] Through the blockchain on-chain governance mechanism, the energy consumption smoothing factor S obtained during the federated learning process is encrypted and distributed; the updated energy consumption smoothing factor S is sent to all edge acquisition and scheduling execution modules in the system to ensure that each module uses the latest parameters in resource scheduling and energy management.

[0057] The on-chain governance mechanism is responsible not only for distributing the energy consumption smoothing factor S, but also for monitoring the real-time synchronization and execution of the energy consumption smoothing factor S by each edge acquisition and scheduling execution module, and automatically verifying the reception and execution status of updated data through smart contracts.

[0058] Compared with the prior art, the beneficial effects of the present invention are:

[0059] This invention addresses the fundamental problem of inaccurate resource demand forecasting due to a lack of detailed data for different classroom types in practical training configurations. It introduces a classroom type identification module, a data acquisition and processing module, a blockchain storage consensus module, an intelligent scheduling and execution module, and a feedback optimization and federated learning module. The system builds a model to subdivide the specific needs of each classroom and uses principal component analysis to establish a classroom type preference matrix. Combining expert experience and user feedback, it quantifies and generates type demand coefficients Dxq to ensure that resource allocation matches actual needs, thus achieving precise resource allocation. Simultaneously, the system collects and processes real-time data points from edge nodes across campuses and uses a multi-chain architecture to maintain metadata for each type of classroom, achieving effective data consensus and intelligent event triggering. This avoids frequent compensations caused by resource scheduling misjudgments and reduces the impact of computing resource overload and frequent power line switching on equipment stability and user experience.

[0060] This invention also employs innovative federated learning technology through feedback optimization and federated learning modules to construct and dynamically update resource prediction and authentication performance models, protecting data privacy while improving overall performance. By using an update formula for the energy consumption smoothing factor S, systematic data optimization and precise adjustment are achieved. The blockchain on-chain governance mechanism ensures the secure encryption of federated learning optimization data and real-time synchronization of edge modules within the system. Simultaneously, smart contracts automatically verify updated status, effectively reducing the misjudgment rate of resource scheduling through scientific management and monitoring. This innovative system overcomes the traditional barriers to multi-campus classroom resource management, enhancing the effectiveness of teaching resources and the stability of equipment, thus improving the overall user experience. Attached Figure Description

[0061] Figure 1 This is a schematic diagram of the overall system framework of the present invention. Detailed Implementation

[0062] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings.

[0063] Many specific details are set forth in the following description in order to provide a full understanding of the invention. However, the invention may also be practiced in other ways different from those described herein, and those skilled in the art can make similar extensions without departing from the spirit of the invention. Therefore, the invention is not limited to the specific embodiments disclosed below.

[0064] Example 1:

[0065] Please see Figure 1 This invention provides a technical solution: a multi-campus collaborative teaching resource blockchain dynamic scheduling system, applied to adjacent campuses sharing high-demand resources, comprising:

[0066] The classroom type identification module is used to create a classroom type preference matrix and, during system initialization and periodic updates, to identify the usage characteristics of classrooms in each campus through the classroom type preference matrix and calculate the corresponding type demand coefficient Dxq.

[0067] The data acquisition and processing module collects real-time environmental status data, computing power data, and power demand data of classrooms at the edge nodes of each campus; it also classifies, segments, and homomorphically encrypts the raw data for different classroom types, and generates data blocks with type labels.

[0068] The blockchain storage consensus module is used to store data blocks marked with each classroom type; for each write operation, an associated smart event is generated and a threshold value M is set to trigger a "data ready" signal in real time; at the same time, a computing power demand fluctuation model is built to analyze the historical fluctuation range of each type of classroom.

[0069] The intelligent scheduling and execution module, upon receiving "data ready," reads the resource configuration parameters of each classroom and the latest fluctuations in computing power demand from the blockchain; simultaneously, it calculates the computing power priority allocation value P and generates a smoothed resource allocation scheme.

[0070] Write the resource allocation plan into the blockchain to trigger the execution of computing power and electricity, and monitor the switching of computing power resources and electricity in real time.

[0071] The feedback optimization and federated learning module sets an optimization cycle and uses federated learning to train the "resource prediction and certification performance model" in parallel on each campus, and calculates the updated energy consumption smoothing factor S. It is then distributed to the edge acquisition and scheduling execution module through on-chain governance to further improve prediction accuracy, reduce the frequency of misjudgments, and reduce the number of power supply switching times.

[0072] The classroom type identification module specifically includes:

[0073] By collecting historical data on classroom usage frequency, usage duration, and user satisfaction in each classroom, principal component analysis is used to extract features from the historical data, including acoustic requirements, equipment requirements, and spatial requirements.

[0074] Based on the extracted features, a classroom type preference matrix is ​​constructed, with different types of classrooms as rows and each feature requirement as a column. Specific feature weights are generated to form the classroom type preference matrix.

[0075] The classroom types include piano rooms, laboratories, and library study rooms; the classroom type preference matrix consists of the acoustic requirements, connectivity requirements, and environmental stability of each type of classroom.

[0076] After identifying the characteristics of classrooms in each campus, the feature weights are multiplied by the actual performance values ​​of the corresponding classrooms, and then the results of all features are summed up to obtain the final type demand coefficient Dxq.

[0077] Furthermore, "acoustic requirements" refer to the classroom's requirements for the sound environment, mainly including indicators such as indoor noise control, reverberation time, and sound insulation performance, in order to ensure the auditory effect of activities such as piano performance and language teaching.

[0078] "Connectivity requirements" refers to the interconnectivity between classrooms and external networks and internal information systems on campus, encompassing parameters such as bandwidth, latency, packet loss rate, and wireless coverage strength, in order to meet the real-time requirements of online experiments, remote teaching, and large-scale data interaction.

[0079] "Environmental stability" refers to the stability level of the classroom in multiple environmental indicators such as temperature, humidity, air quality and lighting, to ensure the reliability of equipment operation and the comfort of teachers and students.

[0080] By taking the above three requirements as the output features of principal component analysis, assigning weights to each, multiplying them by the actual measurement values ​​of each classroom, and summing them up, the type requirement coefficient Dxq of the classroom can be obtained.

[0081] Furthermore, the formula for calculating the type demand factor Dxq is as follows:

[0082]

[0083] Where j represents the total number of features considered in the classroom type preference matrix; The preference weight of classroom type i on feature r is directly taken from the corresponding row and column elements in the classroom type preference matrix constructed after principal component analysis; and i∈{1,2,…,i,…,n} is set, where n represents the total number of classroom types; r∈{1,2,3}, which represent acoustic requirements, equipment requirements and space requirements respectively;

[0084] For classroom type i, the normalized actual performance value on feature r is a quantifiable indicator of acoustic requirements, equipment requirements, and space requirements.

[0085] In this way, the type demand coefficient Dxq provides the system with a quantitative indicator to assess the resource matching needs of different classrooms under specific time and conditions, thereby assisting in making more accurate resource allocation and optimization decisions.

[0086] The data acquisition and processing module includes a data acquisition unit:

[0087] The data acquisition unit collects real-time environmental status data, computing power data, and power demand data of classrooms at the edge nodes of each campus.

[0088] The environmental data includes classroom temperature, humidity, air quality, and lighting intensity;

[0089] Computing power data includes processor utilization, memory usage, and network bandwidth of computer equipment;

[0090] Electricity demand data includes classroom electricity consumption and peak load;

[0091] The collected environmental status data, computing power data, and power demand data are transmitted using an encrypted transmission protocol.

[0092] Furthermore, to meet the specific hardware resource requirements of different classroom types, principal component analysis was used to extract features from the historical usage data of classrooms, namely usage frequency, usage duration, and user satisfaction. The extracted features included acoustic requirements, equipment requirements, and spatial requirements. Based on these features, a classroom type preference matrix was constructed. This matrix uses different types of classrooms as rows, namely piano rooms, laboratories, and library study rooms, and the extracted feature requirements as columns. The matrix was constructed by assigning corresponding weights to each feature based on historical data and expert opinions.

[0093] The data acquisition and processing module is deployed at edge nodes in each campus, and the acquired data is transmitted securely through an encrypted transmission protocol to ensure data integrity and security.

[0094] This real-time data point collection and encrypted transmission mechanism not only protects data privacy but also provides timely and accurate data support for the dynamic scheduling of classroom resources.

[0095] The data acquisition and processing module also includes a threshold stratification processing unit;

[0096] The threshold layering processing unit is used to receive environmental status data, computing power data, and power demand data, and process them through a funnel filter window to filter out noise data.

[0097] Based on different classroom types, threshold criteria for stratification are set, including a first threshold and a second threshold, to stratify the data. Specifically:

[0098] For classroom types, different stratification criteria thresholds are preset at the school-level edge nodes based on expert experience values ​​and historical data medians. Baseline thresholds are calculated based on historical data, including the arithmetic mean and standard deviation of data during the statistical period. Then, the stratification criteria thresholds are obtained by calculation based on the arithmetic mean and standard deviation.

[0099] Furthermore, the specific calculation formula for the threshold of the stratification judgment standard is as follows:

[0100] , where k = 1.2;

[0101] , where k low =1.0;

[0102] The first threshold, The second threshold; and These are the arithmetic mean of the corresponding data during the statistical period. and It is the standard deviation of the corresponding data during the statistical period, k and k low These are the adjustment coefficients for the threshold values ​​used to determine the upper and lower stratification criteria.

[0103] The thresholds for each classroom type are generated and updated in real time. The threshold stratification processing unit continuously calculates the new arithmetic mean and standard deviation and applies an exponentially weighted moving average to smooth the thresholds.

[0104] The real-time data points after funnel filtering are classified into layers. By comparing and evaluating the real-time data points with the layer classification criteria threshold, they are labeled separately, including the first load layer, the second load layer, and the third load layer, thereby generating data blocks with hierarchical labels.

[0105] Furthermore, the settings for the first load tier, the second load tier, and the third load tier are based on statistical analysis of historical data;

[0106] When the real-time data point xt < the first threshold At that time, it is marked as the first load layer T1;

[0107] When the first threshold ≤Real-time data point xt≤Second threshold At that time, it is marked as the second load layer T2;

[0108] When the real-time data point xt > the second threshold When this is done, it is marked as the third load layer T3, thus generating data blocks with hierarchical labels;

[0109] Furthermore, the threshold hierarchical processing unit continuously calculates new arithmetic mean and standard deviation, and applies an exponentially weighted moving average to smooth these thresholds to adapt to dynamic changes in the environment and system operating status. Based on the set thresholds, this labeling process generates data blocks with hierarchical labels, providing accurate data support for subsequent resource scheduling and management.

[0110] The blockchain storage consensus module includes multi-chain architecture units:

[0111] The multi-chain architecture unit is used to set up campus-level edge nodes; an independent sidechain is configured for each classroom type, and a classroom type tag is written in the genesis block of each sidechain;

[0112] For each write operation of the classroom type label, an associated smart event is automatically generated based on the characteristics of the written data; the smart event contains the data's metadata information, namely metadata type, timestamp, data block identifier, and cumulative signature.

[0113] A threshold value M is preset to confirm the number of nodes. When the number of confirmed nodes reaches or exceeds the threshold value M, a "data ready" signal is triggered, and the written data and its associated smart events are encrypted and stored in the corresponding sidechain.

[0114] The blockchain storage consensus module also includes a demand fluctuation model construction unit;

[0115] The demand fluctuation model construction unit collects historical data on resource usage, including the frequency and time of data writing and access for each type of classroom, and inputs this data into the computing power demand fluctuation model.

[0116] Time series analysis is used to predict future data access and write requirements for each type of classroom; the time series model takes into account historical periodic and trend changes to generate a forecast report for each classroom type.

[0117] Based on the prediction results, the resource allocation of each node in the blockchain network is dynamically adjusted; specifically, the support of nodes is automatically increased for the types of classrooms expected during peak demand periods, while the resource allocation is reduced during off-peak demand periods.

[0118] In addition, the demand fluctuation model will monitor in real time the difference between actual data usage and forecasts, and automatically adjust the historical periodic and trend changes for the next monitoring period.

[0119] Furthermore, the detailed operational steps for the multi-chain architecture unit are as follows:

[0120] Each classroom type has its own independent sidechain, and each sidechain writes the corresponding classroom type tag into its genesis block to ensure data separation and dedicated processing. For each write operation of classroom data, associated smart events are automatically generated based on the data characteristics. These smart events record the data's metadata information, including metadata type, timestamp, data block identifier, and cumulative signature, thereby enhancing data security and traceability.

[0121] Establish a threshold value M, which is the value that the number of nodes must reach or exceed in order to trigger the "data ready" signal; when the number of nodes meets this threshold value, the data to be written and its associated smart events will be encrypted and securely stored in the corresponding sidechain;

[0122] Furthermore, the detailed operational steps for constructing the demand fluctuation model unit are as follows:

[0123] Collect historical data, including the frequency of data writing and access for various types of classrooms and the time points, and use this data integration foundation to build a demand fluctuation model;

[0124] A model is built using time series analysis methods. This model will take into account the periodicity and trend changes of historical data to accurately predict future data access and writing needs for each type of classroom.

[0125] Based on the prediction results of the time series model, the resource allocation of each node in the blockchain network is dynamically adjusted. For classroom types with expected high demand periods, more nodes are automatically added to support them; for low demand periods, resource allocation is reduced accordingly. The difference between usage and model predictions is monitored in real time, and the parameters of the prediction model are automatically adjusted based on actual data to enhance the accuracy and adaptability of the model.

[0126] This module, by setting clear operational steps and implementation details, not only improves the system's data processing and storage efficiency, but also significantly enhances resource utilization efficiency and system response speed through intelligent and dynamic management. Compared with existing technologies, this solution has made fundamental improvements and innovations in data security, traceability, and dynamic resource management, demonstrating high practicality and innovation.

[0127] The intelligent scheduling execution module includes a resource scheduling unit;

[0128] After receiving the "data ready" signal transmitted in the blockchain network, the resource scheduling unit instantly accesses the resource configuration parameters of each type of classroom and its latest computing power demand fluctuation information from the blockchain database.

[0129] Based on the obtained resource configuration parameters and the latest computing power demand fluctuation information, the resource demand priority of each type of classroom is calculated using a computing power priority allocation algorithm;

[0130] Furthermore, the computing power priority allocation algorithm calculates the priority allocation value P for the nth classroom type using the following formula:

[0131]

[0132] in, This represents the data access frequency of the nth type of classroom. The time sensitivity index represents the nth type of classroom. This represents the current available computing resources for the nth type of classroom;

[0133] Generate a smoothed resource allocation scheme based on the calculated priority allocation value P to adjust the resource allocation for different types of classrooms.

[0134] The intelligent scheduling and execution module also includes a resource management unit;

[0135] The resource management unit records the resource allocation scheme reconfigured according to the priority allocation value P to the blockchain network, triggering the automatic execution of computing power and power resources;

[0136] During the process, it is controlled by a pre-defined smart contract, which simultaneously monitors and adjusts the switching between computing resources and electricity in real time.

[0137] Furthermore, the resource management unit first responds to the "data ready" signal from the blockchain network, meaning that there is new data or events that need to be processed; then, the resource scheduling unit accesses the resource configuration parameters of all classroom types and their latest computing power demand fluctuation information from the blockchain database; this information includes data access frequency, time sensitivity, and the current availability of computing resources;

[0138] Using a computing power priority allocation algorithm, the unit calculates the resource demand priority for each classroom type; the calculation of the priority allocation value P for the nth classroom type involves three main parameters: data access frequency, time sensitivity index, and the current availability of computing resources;

[0139] Among them, data access frequency reflects how often data for a specific classroom type is queried or used, which helps identify which classrooms have more frequent resource demands.

[0140] Time sensitivity metrics are used to measure the rigor of time response for classroom types that require real-time or near-real-time processing, including laboratory control systems.

[0141] The current availability of computing resources indicates the amount of computing resources currently available for this type of classroom, which determines the upper limit of resources that can be allocated to this classroom;

[0142] A smoothed resource allocation scheme is generated based on the calculated priority allocation value P, aiming to optimize the overall resource utilization efficiency and response speed; the specific process is as follows:

[0143] A first resource demand priority threshold P1 and a second resource demand priority threshold P2 are preset, with the first resource demand priority threshold P1 being greater than the second resource demand priority threshold P2. These are then compared and evaluated with the priority allocation value P. The specific details are as follows:

[0144] When the priority allocation value P is greater than or equal to the first resource demand priority threshold P1, the current classroom is marked as a high-priority classroom. These classrooms will receive additional computing resources, storage space, and network bandwidth to ensure that they can handle high-frequency data access and time-sensitive tasks. At this time, the resource allocation scheme includes increasing server quotas, increasing network priority, or allocating more memory and CPU resources.

[0145] When the first resource demand priority threshold P1 > the priority allocation value P ≥ the second resource demand priority threshold P2, the current classroom is marked as a medium-priority classroom. At this time, the current resource configuration remains unchanged to ensure the stable operation of the classroom without investing too many unnecessary resources. The performance and resource requirements of these classrooms are regularly evaluated to prevent future changes in demand.

[0146] When the second resource demand priority threshold P2 is greater than the priority allocation value P, the current classroom is marked as a low-priority classroom. At this time, resource allocation is reduced, including reducing network bandwidth or reducing the server resources allocated to it. At the same time, the released resources are reallocated to high-priority classrooms, thereby optimizing the overall resource utilization efficiency.

[0147] Implementing the resource allocation plan into actual operation includes adjusting server, storage, and bandwidth allocations; the resource management unit needs to regularly monitor the implementation effect and adjust the weight values ​​and re-evaluate the priority allocation value P based on actual operating data;

[0148] Based on the priority allocation scheme generated by the resource scheduling unit, the resource management unit records the new resource allocation configuration in the blockchain network; this step ensures that any resource adjustment is immutable and traceable.

[0149] The resource management unit triggers automatic execution operations of computing power and power resources. This operation is controlled by a pre-defined smart contract to ensure that the execution process is automated and conforms to preset rules.

[0150] In addition, the unit will monitor and adjust the switching between computing resources and power in real time to adapt to sudden changes in demand or optimize resource utilization efficiency.

[0151] The Feedback Optimization and Federated Learning module specifically includes:

[0152] The configuration and launch of federated learning are used to jointly optimize and train resource prediction and certification performance models without sharing campus data, thereby protecting data privacy while improving the global performance of the model.

[0153] The energy consumption smoothing factor S is calculated and updated using the federated learning results, employing the following update formula:

[0154]

[0155] in, α represents the smoothing parameter, where 0 < α < 1, Sold is the previous energy consumption smoothing factor, and Slear is the updated value derived from the current federated learning results.

[0156] The Feedback Optimization and Federated Learning module also includes:

[0157] Through the blockchain on-chain governance mechanism, the energy consumption smoothing factor S obtained during the federated learning process is encrypted and distributed; the updated energy consumption smoothing factor S is sent to all edge acquisition and scheduling execution modules in the system to ensure that each module uses the latest parameters in resource scheduling and energy management.

[0158] The on-chain governance mechanism is responsible not only for distributing the energy consumption smoothing factor S, but also for monitoring the real-time synchronization and execution of the energy consumption smoothing factor S by each edge acquisition and scheduling execution module, and automatically verifying the reception and execution status of updated data through smart contracts.

[0159] Furthermore, local training nodes are deployed in each campus, with each node running a local model; these models are configured to be trained for specific prediction and resource allocation needs, namely energy consumption prediction.

[0160] Encryption techniques, including homomorphic encryption, are used to ensure the security and privacy of data during transmission.

[0161] Initiate a federated learning process, in which nodes do not share actual data, but instead exchange model parameters or learned features;

[0162] By using blockchain technology, the updated energy consumption smoothing factor S is encrypted and distributed to all relevant edge acquisition and scheduling execution modules, which ensures that the transmission of the update is secure and tamper-proof.

[0163] The smart contract is configured to automatically detect and verify the reception and execution status of each module in response to the new energy consumption smoothing factor S, ensuring the consistency and up-to-date status of the entire system.

[0164] This design not only ensures data privacy and the effectiveness of model training, but also enhances the transparency and security of system management through the application of smart contracts and blockchain. Compared with existing technologies, this solution demonstrates significant improvements in protecting personal data privacy, optimizing resource allocation strategies, and enhancing the reliability and security of distributed system management. This integrated approach utilizing federated learning and blockchain technologies provides a novel and practical solution for energy management and resource allocation in modern educational institutions.

[0165] Example 2:

[0166] An educational institution with three types of classrooms, including piano rooms, laboratories, and library study rooms, is located across three different campuses. The main purpose of the system is to optimize resource allocation and energy management for these classrooms.

[0167] 1. Classroom type preference matrix:

[0168] Table 1 shows the preference weight parameters for different classroom types based on classroom usage frequency, usage duration, and user satisfaction.

[0169]

[0170] 2. Resource scheduling and energy management:

[0171] Based on the federated learning results, assuming the updated energy consumption smoothing factor S is calculated using the following formula and example values:

[0172]

[0173] Where Sold=0.5; Slear=0.65; Snew=0.7×0.5+0.3×0.65=0.545;

[0174] 3. Blockchain storage and data readiness signals:

[0175] After processing through a multi-chain structure, assuming that the data readiness signal for the piano room has been triggered, this will activate the intelligent scheduling system for resource allocation and energy management.

[0176] It should be noted that all calculation formulas in this application employ regression analysis, including but not limited to machine learning algorithms, to deeply analyze the collected parameters and identify their natural trends and interrelationships. Specialized software, such as Python's Scikit-learn library or the R language, is used to automatically generate mathematical models that match the data. Then, cross-validation and other methods are used to objectively evaluate the model performance, and continuous feedback and optimization are combined to ensure that the created formulas truly reflect the inherent laws of the data, thereby guaranteeing their effectiveness and accuracy. In all calculation formulas in this application, the parameters in each formula undergo dimensionless processing within a consistent range to ensure that different physical quantities are compared on the same scale; dimensionless processing techniques include, but are not limited to, min-max-normalization and Z-score standardization.

[0177] The technical solution of this invention, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as a computer floppy disk, read-only memory (ROM), random-access memory (RAM), flash memory, hard disk, or optical disk, etc., including several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods of various embodiments of this invention.

[0178] The logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as a sequenced list of executable instructions for implementing logical functions, and can be embodied in any computer-readable medium for use by, or in conjunction with, an instruction execution system, apparatus, or device (such as a computer-based system, a processor-including system, or other system that can fetch and execute instructions from, an instruction execution system, apparatus, or device). For the purposes of this specification, "computer-readable medium" can be any means that can contain, store, communicate, propagate, or transmit programs for use by, or in conjunction with, an instruction execution system, apparatus, or device.

[0179] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

Claims

1. A multi-campus collaborative teaching resource blockchain dynamic scheduling system, applied to adjacent campuses sharing high-demand resources, characterized in that... include: The classroom type identification module collects historical multidimensional status data for each classroom and extracts features from this data using principal component analysis, including acoustic requirements, equipment requirements, and spatial requirements. Based on these extracted features, a classroom type preference matrix is ​​constructed, with different classroom types as rows and each feature requirement as a column, generating specific feature weights to form the classroom type preference matrix. This matrix consists of the acoustic requirements, connectivity requirements, and environmental stability of each classroom type. During system initialization and periodic updates, the module identifies the usage characteristics of classrooms in each campus using the classroom type preference matrix, multiplies the feature weights by the actual performance values ​​of the corresponding classrooms, and then sums up the results of all features to obtain the final type requirement coefficient Dxq. The data acquisition and processing module collects real-time environmental status data, computing power data, and power demand data of classrooms at the edge nodes of each campus; it also classifies, segments, and homomorphically encrypts the raw data for different classroom types, and generates data blocks with type labels. The blockchain storage consensus module is used to store data blocks marked with each classroom type; for each write operation, an associated smart event is generated and a threshold value M is set to trigger a "data ready" signal in real time; at the same time, a computing power demand fluctuation model is built to analyze the historical fluctuation range of each type of classroom. The intelligent scheduling and execution module, upon receiving the "data ready" signal, reads the resource configuration parameters and the latest fluctuations in computing power demand for each classroom in the blockchain; at the same time, it calculates the computing power priority allocation value P and generates a smoothed resource allocation scheme. Write the resource allocation plan into the blockchain to trigger the execution of computing power and electricity, and monitor the switching of computing power resources and electricity in real time. The feedback optimization and federated learning module sets an optimization cycle and uses federated learning to train the "resource prediction and certification performance model" in parallel on each campus, and calculates the updated energy consumption smoothing factor S. It is then distributed to the edge acquisition and scheduling execution module through on-chain governance to further improve prediction accuracy, reduce the frequency of misjudgments, and reduce the number of power supply switching times.

2. The multi-campus collaborative teaching resource blockchain dynamic scheduling system according to claim 1, characterized in that: The data acquisition and processing module includes a data acquisition unit: The data acquisition unit is used to collect environmental status data, operational computing power data, and power demand data. The environmental data includes classroom temperature, humidity, air quality, and lighting intensity; Computing power data includes processor utilization, memory usage, and network bandwidth of computer equipment; Electricity demand data includes classroom electricity consumption and peak load; The collected environmental status data, computing power data, and power demand data are transmitted using an encrypted transmission protocol.

3. The multi-campus collaborative teaching resource blockchain dynamic scheduling system according to claim 2, characterized in that: The data acquisition and processing module also includes a threshold stratification processing unit; The threshold layering processing unit is used to receive environmental status data, computing power data, and power demand data, and process them through a funnel filter window to filter out noise data. Based on different classroom types, threshold criteria for stratification are set, including a first threshold and a second threshold, to stratify the data. Specifically: For different classroom types, different stratification criteria thresholds are preset at the campus-level edge nodes; baseline thresholds are calculated based on historical data, including the arithmetic mean and standard deviation of the data during the statistical period. The thresholds for each classroom type are generated and updated in real time. The threshold stratification processing unit continuously calculates the new arithmetic mean and standard deviation and applies an exponentially weighted moving average to smooth the thresholds. The real-time data points after funnel filtering are classified into layers. By comparing and evaluating the real-time data points with the layer classification criteria threshold, they are labeled separately, including the first load layer, the second load layer, and the third load layer, thereby generating data blocks with hierarchical labels.

4. The multi-campus collaborative teaching resource blockchain dynamic scheduling system according to claim 3, characterized in that: The blockchain storage consensus module includes multi-chain architecture units: The multi-chain architecture unit is used to set up campus-level edge nodes; an independent sidechain is configured for each classroom type, and a classroom type tag is written in the genesis block of each sidechain; For each write operation of the classroom type label, an associated smart event is automatically generated based on the characteristics of the written data; the smart event contains the data's metadata information, namely metadata type, timestamp, data block identifier, and cumulative signature. A threshold value M is preset to confirm the number of nodes. When the number of confirmed nodes reaches or exceeds the threshold value M, a "data ready" signal is triggered, and the written data and its associated smart events are encrypted and stored in the corresponding sidechain.

5. The multi-campus collaborative teaching resource blockchain dynamic scheduling system according to claim 4, characterized in that: The blockchain storage consensus module also includes a demand fluctuation model construction unit; The demand fluctuation model construction unit collects historical data on resource usage, including the frequency and time of data writing and access for each type of classroom, and inputs this data into the computing power demand fluctuation model. Time series analysis is used to predict future data access and write requirements for each type of classroom; the time series model takes into account historical periodic and trend changes to generate a forecast report for each classroom type. Based on the prediction results, the resource allocation of each node in the blockchain network is dynamically adjusted; specifically, the support of nodes is automatically increased for the types of classrooms expected during peak demand periods, while the resource allocation is reduced during off-peak demand periods. In addition, the demand fluctuation model will monitor in real time the difference between actual data usage and forecasts, and automatically adjust the historical periodic and trend changes for the next monitoring period.

6. The multi-campus collaborative teaching resource blockchain dynamic scheduling system according to claim 5, characterized in that: The intelligent scheduling execution module includes a resource scheduling unit; After receiving the "data ready" signal transmitted in the blockchain network, the resource scheduling unit instantly accesses the resource configuration parameters of each type of classroom and its latest computing power demand fluctuation information from the blockchain database. Based on the obtained resource configuration parameters and the latest computing power demand fluctuation information, the resource demand priority of each type of classroom is calculated using a computing power priority allocation algorithm; Generate a smoothed resource allocation scheme based on the calculated priority allocation value P to adjust the resource allocation for different types of classrooms.

7. The multi-campus collaborative teaching resource blockchain dynamic scheduling system according to claim 6, characterized in that: The intelligent scheduling and execution module also includes a resource management unit; The resource management unit records the resource allocation scheme reconfigured according to the priority allocation value P to the blockchain network, triggering the automatic execution of computing power and power resources; During the process, it is controlled by a pre-defined smart contract, which simultaneously monitors and adjusts the switching between computing resources and electricity in real time.

8. The multi-campus collaborative teaching resource blockchain dynamic scheduling system according to claim 7, characterized in that: The Feedback Optimization and Federated Learning module specifically includes: The configuration and launch of federated learning are used to jointly optimize and train resource prediction and certification performance models without sharing campus data, thereby protecting data privacy while improving the global performance of the model. The energy consumption smoothing factor S is calculated and updated using the federated learning results, employing the following update formula: ； in, α represents the smoothing parameter, where 0 < α < 1, Sold is the previous energy consumption smoothing factor, and Slear is the updated value derived from the current federated learning results.

9. The multi-campus collaborative teaching resource blockchain dynamic scheduling system according to claim 8, characterized in that: The Feedback Optimization and Federated Learning module also includes: Through the blockchain on-chain governance mechanism, the energy consumption smoothing factor S obtained during the federated learning process is encrypted and distributed; the updated energy consumption smoothing factor S is sent to all edge acquisition and scheduling execution modules in the system to ensure that each module uses the latest parameters in resource scheduling and energy management. The on-chain governance mechanism is responsible not only for distributing the energy consumption smoothing factor S, but also for monitoring the real-time synchronization and execution of the energy consumption smoothing factor S by each edge acquisition and scheduling execution module, and automatically verifying the reception and execution status of updated data through smart contracts.

Images