Multi-module collaborative management vocational education system based on edge computing

Abstract

The invention provides a vocational education system for multi-module collaborative management based on edge computing, and relates to the field of education management. Comprising a dynamic task state machine module for monitoring multi-dimensional parameters such as task progress, network delay, equipment load and security risk in real time and triggering a permission migration instruction; the authority fragmentation edge cache module is used for disassembling the authority into atomization units (APU) and storing the APUs in edge nodes in a blocking manner, and verifying the integrity through a Hash authority fingerprint chain; the conflict resolution protocol module dynamically generates a priority weight based on task remaining time, records an operation log chain and executes local rollback; the edge node cluster module dynamically switches calculation, routing and storage roles to optimize resource allocation; the system adapts to different security level tasks through a dynamic blocking strategy and a recombination threshold value, priority preemption and parallel transmission are adopted to reduce delay, and an encryption timestamp and a forward hash pointer are utilized to guarantee operation traceability.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of education management, in particular to a multi-module collaborative management vocational education system based on edge computing. BACKGROUND

[0002] With the acceleration of digital transformation of vocational education, real-time interaction and resource collaboration demand is increasing; edge computing, with its low latency and distributed characteristics, has become a key technology for optimizing the efficiency of education systems; multi-terminal device access and complex task scenarios require higher requirements for dynamic resource scheduling, permission management and conflict resolution, and it is urgent to build an efficient and reliable technical framework to support intelligent teaching and management needs.

[0003] The current vocational education system mostly uses a centralized cloud computing architecture, which allocates tasks and permissions through a centralized server; permission management is usually based on static role division, and a fixed storage strategy is used to ensure data security; conflict resolution relies on manual intervention or simple priority queues, and task scheduling is mainly based on preset rules; the existing solution relies on a stable network environment and reduces the risk of node failure through redundant backup, but flexibility is limited.

[0004] The existing technology has the following shortcomings: the centralized architecture results in high task response delay, making it difficult to meet real-time requirements; static permission management cannot adapt to dynamic scenarios, and permission migration efficiency is low; the conflict resolution mechanism lacks adaptive ability, and resource preemption and operation conflicts are easily generated when multiple tasks are concurrent; the fixed node role leads to uneven load, and the system stability decreases when the network fluctuates, making it difficult to support the flexible needs of complex education scenarios. SUMMARY

[0005] (I) Technical problems solved

[0006] In view of the shortcomings of the prior art, the present application provides a multi-module collaborative management vocational education system based on edge computing to solve the problems of high task response delay caused by the centralized architecture, low permission migration efficiency of static permission management that cannot adapt to dynamic scenarios, resource preemption and operation conflicts caused by the lack of adaptive ability of the conflict resolution mechanism when multiple tasks are concurrent, and the uneven load and insufficient system stability caused by the fixed node role under network fluctuations.

[0007] (II) Technical solutions

[0008] To achieve the above purpose, the present application realizes the following technical solutions: a multi-module collaborative management vocational education system based on edge computing, comprising: a dynamic task state machine module for real-time monitoring of multi-dimensional state parameters such as task progress, network delay, device load, security risk level, data integrity and time sensitivity, and triggering a permission migration instruction according to the comprehensive evaluation results of the multi-dimensional state parameters;

[0009] The permission fragmentation edge caching module decomposes permissions into atomic permission units (APUs) and divides them into several data blocks. These blocks are stored in the computing unit, routing unit, and storage unit of the edge node cluster module according to the block division strategy corresponding to the preset security level. This generates a hash permission fingerprint chain formed by concatenating the block hash values ​​in the storage order to verify the integrity of permissions.

[0010] The conflict resolution protocol module dynamically generates priority weights based on the remaining time of the task, preempts the right to perform operations based on the priority weights, and records the timestamps of conflict operations, user identifiers and permission unit statuses through the operation log chain, and performs partial rollback.

[0011] The edge node cluster module includes a computing unit, a routing unit, and a storage unit that dynamically switch roles, and adjusts node roles and permission storage paths according to the instructions of the dynamic task state machine module.

[0012] Preferably, the dynamic task state machine module collects multi-dimensional state parameters such as task progress, network latency, device load, security risk level, data integrity, and time sensitivity in real time through sensors and agent programs deployed on edge computing nodes; the task progress is calculated by obtaining the ratio of the currently completed steps to the total steps through the task management interface; the network latency is measured by the round-trip time (RTT) of probe messages between edge nodes; the device load is obtained by weighted summation of CPU utilization, memory usage, and disk I / O rate of the computing unit, routing unit, and storage unit; and the security risk level is fed back by the permission fragmentation edge caching module. The data integrity verification result is compared with a preset risk threshold to determine the data integrity; the data integrity is calculated by statistically analyzing the complete block ratio based on the block-by-block verification results of the hash permission fingerprint chain; the time sensitivity is dynamically adjusted based on the difference between the task deadline and the current time; the module inputs the above parameters into a comprehensive evaluation model to generate a normalized status score; when the score exceeds a preset threshold, a permission migration instruction is triggered, and the specific migration target node is dynamically selected by the edge node cluster module based on network latency and load balancing algorithms; during the migration process, the module freezes the permission operations of the original node, and unfreezes it after the target node completes APU block data reception and integrity verification.

[0013] Preferably, the permission fragmentation edge cache module disassembles user permissions into atomic permission units (APUs), each of which is the smallest unit of permission operation; when disassembling, a task security level defined by the dynamic task state machine module is selected to select a block strategy: the APUs of a high-risk task are split into N data blocks according to a first block number, and the APUs of a low-risk task are split into M data blocks according to a second block number, where N>M; after the splitting, a dynamic index serial number is attached to each data block, and the data blocks are distributed to the computing unit, routing unit and storage unit of the edge node cluster module according to a storage rule corresponding to a preset security level; the data blocks of the high-risk task adopt a redundant storage strategy and are stored in at least three nodes; the data blocks of the low-risk task adopt a single-copy storage; after the storage is completed, the module calculates the hash values of the blocks in the block order and concatenates to generate a hash permission fingerprint chain; the hash permission fingerprint chain is stored in a block chain node to ensure that it cannot be tampered with; when verifying the permission, the hash values of the blocks need to be checked in a chained order, and if a reorganization threshold is continuously met, i.e., all N blocks of the high-risk task and at least M-1 blocks of the low-risk task, it is determined that the permission is complete.

[0014] Preferably, the conflict resolution protocol module dynamically generates a priority weight according to the remaining time of a task; the remaining time is calculated by the difference between the task deadline timestamp and the current system timestamp, and is rounded to the nearest hour; the priority weight is inversely proportional to the remaining time, and the calculation formula is: weight = 1 / (remaining time + 1); when multiple task requests the same APU, the module sorts the tasks in descending order of weight, and the high-weight task forcibly interrupts the permission operation of the low-weight task, and the operation request of the interrupted task is added to the pending queue; after the preemption operation is triggered, the module records the timestamp, user ID, APU identifier, operation type and operation result of the conflict operation to an operation log chain; the operation log chain adopts a chained structure, and each operation record contains a forward hash pointer and an encrypted timestamp, the forward hash pointer is generated by calculating the current block hash value and the previous record pointer value, and the encrypted timestamp adopts RSA asymmetric encryption; when rolling back, the conflict resolution protocol module decrypts the timestamp in reverse order of time and verifies the continuity of the forward hash pointer; if a hash mismatch or decryption failure is detected, the rollback is terminated and a security alarm is triggered.

[0015] Preferably, the preset delay threshold, the first load threshold, the second load threshold and the third load threshold are set based on historical data training and dynamic adjustment mechanism in the system initialization stage; the first load threshold is set as 70% of the safe upper limit of CPU utilization of the computing unit, when the device load is lower than the value, the edge collaborative mode is allowed to be triggered, and the APU block is migrated from the high delay node to the low delay node, wherein the safe upper limit of CPU utilization includes: the default of the computing unit is 80%, the default of the routing unit is 70%, and the default of the storage unit is 60%; the second load threshold is set as 90% of the maximum safe load value, when the device load exceeds the value, the resource protection state is forced to enter, and the non-critical APU is frozen and its write operation is limited; the maximum safe load value is determined by stress test and historical data analysis in the system initialization stage, which represents the upper limit of the load that the device can stably run, and its calculation formula is: maximum safe load value = α˙CPU+β˙memory+γ˙disk I / O, and the weight coefficient is set according to the node type; the third load threshold is set as 85% of the CPU utilization of the computing unit, when the computing unit load exceeds the value, its role is dynamically switched to the routing unit to optimize the communication path; when the system is deployed, the average network delay, device load peak and task execution success rate data of the edge node are collected by simulating the task scene; the preset delay threshold is 1.5 times of the historical average delay, if the actual delay exceeds the value, it is determined as high delay state; the above threshold is dynamically fine-tuned according to the node performance fluctuation and task type during operation, for example, the delay threshold is automatically reduced by 10% in high-risk task scene to improve the response speed.

[0016] Preferably, the edge node cluster module includes a computing unit, a routing unit and a storage unit, and the roles of each unit are dynamically switched according to the instructions of the dynamic task state machine module; when the CPU load of the computing unit exceeds the third load threshold, it is switched to the routing unit, which is responsible for optimizing the transmission path of the APU block data; the routing unit calculates the shortest delay path based on the network topology data, and dynamically adjusts the block transmission order; when path congestion is detected, the routing unit splits the block data into sub-streams according to the dynamic index serial number, and transmits them in parallel through multiple low-delay paths; the storage unit recombines the blocks according to the serial number, and initiates a retransmission request to the sending node if missing or out-of-order is detected; if the retransmission times out, the authority migration instruction is triggered; the storage unit dynamically allocates paths according to the storage requirements of the APU block data, and starts the standby node takeover process when the node is disconnected, to ensure data availability.

[0017] Preferably, the generation of the shortest delay path is based on the network topology data collected by the routing unit of the edge node cluster module in real time; the routing unit periodically sends probe packets to adjacent nodes, measures the round-trip time (RTT) between nodes and constructs a delay matrix; based on the delay matrix, the routing unit calculates the minimum cumulative delay path from the source node to the target node using a dynamic programming algorithm, and preferentially selects links with fewer hops and stable delay during path selection; if multiple paths have the same delay, the path with the lowest load is selected according to the node load balancing strategy; during path calculation, the routing unit excludes nodes with a delay exceeding a preset delay threshold to ensure path availability; the calculation result is updated to the routing table in real time for calling by the APU when transmitting the data blocks.

[0018] Preferably, the first and second recombination threshold values are strictly bound to the task safety level; for high-risk tasks, the first recombination threshold value requires that all N blocks pass the hash check to ensure the integrity of the authority; for low-risk tasks, the second recombination threshold value allows at most 1 block to fail the check, i.e., M-1 / M, to balance efficiency and security; when the dynamic task state machine module detects that the network environment has deteriorated, such as a delay fluctuation rate exceeding 20%, the recombination threshold value for low-risk tasks is automatically raised to M / M to temporarily enhance the strictness of the check; after the task is completed, the threshold value returns to the initial setting.

[0019] Preferably, when the authority fragmentation edge cache module triggers the authority migration instruction, the edge node cluster module selects the target node according to the calculation result of the shortest delay path; if the load of the target node exceeds the first load threshold value, the suboptimal node is selected in ascending order of load; during the migration process, the APU block data is transmitted through the shortest delay path, and if the path is interrupted, the routing unit immediately switches to the backup path, i.e., the path with the second lowest delay; the switching time of the backup path must be less than a preset timeout threshold value of 500 ms, otherwise the migration is determined to have failed and a second retry is triggered; after the migration is completed, the target node sends a confirmation signal to the dynamic task state machine module to update the node index information in the hash authority fingerprint chain

[0020] Preferably, when the dynamic task state machine module detects that the network delay exceeds a preset delay threshold value and the device load is lower than the first load threshold value, if the current task safety level is high-risk, the edge coordination mode is triggered; the module selects the edge node with the lowest delay as the target node and migrates the APU block data from the high-delay node to the target node; before the migration, the module freezes the write operation of the original node and checks the storage capacity and load state of the target node; during the migration process, the block data is transmitted through an encrypted channel, and the target node performs hash check immediately after receiving the data; if the check passes, the hash authority fingerprint chain is updated and the freeze is removed; if the check fails, a new node is selected and a second migration is triggered; after the migration is completed, the module synchronizes the task state to the conflict resolution protocol module to adjust the priority weight.

[0021] Preferably, for low-risk tasks, the permission fragmentation edge cache module adopts a second number of blocks (M) and a second reorganization threshold (M-1); when a user initiates an APU access request, the nearest edge node, i.e. the node with the lowest delay to the user terminal, collects M blocks stored in the computing unit, the routing unit and the storage unit; the node calculates the local hash value of each block and compares it with the hash value of the corresponding serial number in the hash permission fingerprint chain; if M-1 consecutive blocks pass the verification, the complete APU is reorganized and access permission is granted; if the verification fails, a permission migration instruction is triggered; during migration, the module preferentially selects nodes with a load lower than the second load threshold and adopts a single-copy storage strategy to reduce overhead; after reorganization is complete, the module feeds back the results to the dynamic task state machine module to update the task progress.

[0022] Preferably, in the operation log chain generated by the conflict resolution protocol module, each operation record contains a forward hash pointer and an encrypted timestamp; the forward hash pointer is generated by calculating the current block hash value and the previous record pointer value using the SHA-256 algorithm; the encrypted timestamp is encrypted using the RSA algorithm on the operation timestamp, and the private key is held by the system administrator; during rollback, the module decrypts the timestamp in reverse order of time and verifies the continuity of the forward hash pointer piece by piece; if a record fails to decrypt or the hash value does not match, it is determined that the log chain has been tampered with, and rollback is immediately terminated and a security alert is triggered; the alert information is broadcast to all relevant nodes through the routing unit of the edge node cluster module, freezing the operation permissions of the affected APU until manual intervention is complete.

[0023] Preferably, when multiple trigger conditions are met simultaneously, the dynamic task state machine module executes the resource protection state with a higher priority than the edge collaboration mode; when the device load exceeds the second load threshold, the module immediately triggers the resource protection state, freezing non-critical APUs such as data queries and material upload write operations, and limiting their block migration; at the same time, the dynamic task state machine module releases part of the computing unit resources and switches them to routing units to optimize the communication path; if the network delay still exceeds the threshold at this time and the task is high-risk, after the resource protection state takes effect, the module selectively executes the edge collaboration mode according to the remaining resource capacity; when the task progress reaches the preset threshold and the permission integrity verification passes, the module switches to the task completion state, releasing all APUs and cleaning up the temporary storage blocks.

[0024] Preferably, when the routing unit detects path congestion, it splits the APU chunked data into K sub-streams according to the dynamic index sequence number, K being the number of available low-delay paths; each sub-stream contains a group of consecutive chunk sequence numbers, for example, sub-stream 1 contains sequence numbers 1-10, and sub-stream 2 contains sequence numbers 11-20; after splitting, the sub-streams are transmitted in parallel through different paths; the storage unit reassembles the data according to the chunk sequence number, and if a sequence number gap is detected, requests the missing chunk from the sending node; if the sending node does not receive a retransmission request response within a preset time, it determines that the path is invalid, triggering the authority migration instruction; during migration, the module reassigns the unsuccessfully transmitted chunks to other available paths and updates the storage node information in the hash authority fingerprint chain.

[0025] Preferably, the dynamic task state machine module synchronizes task progress parameters such as the number of completed chunks, remaining time, and current score to the conflict resolution protocol module and the authority fragmentation edge cache module after each state transition; the conflict resolution protocol module recalculates the priority weight based on the updated progress parameters and adjusts the task order in the pending queue; the authority fragmentation edge cache module dynamically optimizes the chunk storage strategy based on the progress parameters, for example, allocating more redundant chunks to high-risk tasks as the deadline approaches; at the same time, the authority integrity verification result is fed back to the dynamic task state machine module in real time for safety risk level correction and state transition triggering; the above feedback loop ensures that the system achieves dynamic load balancing and safety control under the cooperation of multiple modules.

[0026] (Three) beneficial effects

[0027] The present application provides a multi-module collaborative management of vocational education system based on edge computing. It has the following beneficial effects:

[0028] 1. The present application uses a dynamic task state machine to monitor multi-dimensional state parameters in real time and trigger authority migration instructions, significantly reducing network delay and device load; the authority fragmentation edge cache module uses a hash authority fingerprint chain to verify each chunk, ensuring the integrity of the authority and the controllability of safety risks; the conflict resolution protocol dynamically adjusts the priority weight based on the remaining time of the task, enabling seamless preemption and local rollback of high-sensitivity tasks, and improving system response efficiency and resource utilization.

[0029] 2. In the present application, the edge node cluster module dynamically switches roles and optimizes communication paths, combining chunk sub-stream parallel transmission and missing retransmission mechanisms to effectively alleviate network congestion; authority chunks are redundantly stored according to safety levels, and the reassembly threshold is differentiated, balancing the data reliability of high-risk tasks and the fast access of low-risk tasks; the operation log chain is encrypted and verified, and the rollback mechanism prevents tampering, ensuring system consistency and real-time safety alerts under multi-user concurrent operation. DETAILED DESCRIPTION

[0030] The technical solutions in the embodiments of the present application will be clearly and completely described below. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all the other embodiments obtained by those skilled in the art without creative work fall within the protection scope of the present application.

[0031] Embodiment one:

[0032] The embodiment of the application provides a multi-module cooperative management professional education system based on edge computing, which comprises the following steps: when a user initiates a high-risk task involving fund approval in a professional education system, a dynamic task state machine module collects multi-dimensional state parameters of task progress, network delay, device load, security risk level, data integrity and time sensitivity in real time through sensors and agent programs deployed on edge nodes; the task progress is calculated as 60% by obtaining the ratio of the current completed steps to the total steps through a task management interface; the network delay is measured as 220 milliseconds by the round-trip time of probe packets between edge nodes; the device load is obtained by weighted summation of CPU utilization rate 75%, memory occupancy rate 65% and disk I / O rate 50 megabytes per second of the calculation unit, routing unit and storage unit, and the load value is 68%; the security risk level is determined as high risk by comparing the data integrity check result fed back by the permission fragmentation edge cache module with the preset risk threshold; the data integrity is calculated as 100% by the proportion of complete blocks through block-by-block check result statistics of the hash permission fingerprint chain; the time sensitivity is dynamically adjusted to a high sensitive level according to the difference between the task deadline and the current time, which is 6 hours; the dynamic task state machine module inputs the above parameters into a comprehensive evaluation model to generate a normalized state score of 85, which exceeds the preset threshold of 80, and triggers a permission migration instruction; the migration target node is dynamically selected by the edge node cluster module according to the network delay and the load balancing algorithm, and the node B with the lowest delay of 150 milliseconds and the load of 55% of the node B; during the migration process, the dynamic task state machine module freezes the permission operation of the original node A; the permission fragmentation edge cache module disassembles the fund approval permission into atomic permission units, splits the first block number of the high-risk task into 5 data blocks, adds dynamic index numbers 1 to 5 to each data block, and distributes them to the calculation unit, routing unit and storage unit of the edge node cluster module according to the redundant storage strategy, and each block is stored in at least three nodes; after the storage is completed, the module calculates the hash value of each block in the block order and generates a hash permission fingerprint chain by concatenation, and stores it in a blockchain node; when verifying the permission, the nearest node B of the edge node cluster module collects the 5 block data stored in each unit, calculates the local hash value of each block and compares it with the hash permission fingerprint chain block by block, and if the continuous 5 blocks are all checked, the complete atomic permission unit is reorganized and the access permission is granted; the conflict resolution protocol module dynamically generates a priority weight of 1 / 7, which is equal to 0.14, according to the remaining time 6 hours of the task, and when another low-priority task weight 0.When the same atomic permission unit is requested, the high-weight task forcibly interrupts the operation of the low-weight task, adds it to the pending queue, and records the timestamp, user identifier, and permission unit state of the conflicting operation through the operation log chain. During rollback, the module reverses the operation log chain to the state of the last operation that passed the data integrity check and re-executes the check process. The edge node cluster module detects the disconnection of node C during migration and immediately triggers the standby node D to take over the partitioned data of the storage unit, ensuring data availability. After the task progress reaches the preset threshold of 100% and the permission integrity check passes, the dynamic task state machine module switches to the task completion state, releases all atomic permission units, and cleans up the temporary storage partitions.

[0033] Embodiment Two:

[0034] The difference between this embodiment and embodiment one is that this embodiment takes a low-risk task uploaded by a user as an example. When the dynamic task state machine module collects the task progress as 30%, the network delay as 180 milliseconds, the device load as the weighted sum of CPU utilization rate 60%, memory occupancy rate 40%, and disk I / O rate 30 megabytes per second, the security risk level is determined as low risk, the data integrity complete block proportion is 95%, and the time sensitivity is adjusted as low sensitive level according to the remaining time of the task 48 hours; the comprehensive evaluation model generates a state score of 65, which is lower than the migration threshold 80, and does not trigger the permission migration instruction; the permission fragmentation edge cache module disassembles the material upload permission into atomized permission units, splits it into 3 data blocks according to the second block number of the low-risk task, and allocates it to the storage unit of the edge node cluster module using a single-copy storage strategy, which is stored in only one node; after the generation of the hash permission fingerprint chain, the nearest edge node E collects 3 block data during permission verification, calculates the local hash value and compares it with the corresponding serial number in the chain, and if 2 consecutive blocks pass the verification, the complete atomized permission unit is reassembled; when the user initiates an access request, the conflict resolution protocol module calculates the priority weight as 1 / 49 approximately equal to 0.02 according to the remaining time 48 hours, and if it conflicts with other low-weight tasks, it is processed according to the queue order without triggering the preemption operation; when the operation log chain records the operation information, only a lightweight encryption strategy is used; when the edge node cluster module detects path congestion, the routing unit splits the block data into 2 sub-streams according to the dynamic index serial number, sub-stream 1 contains serial numbers 1 to 2, and sub-stream 2 contains serial number 3, which are transmitted in parallel through two low-delay paths; when the storage unit reassembles according to the serial number, if block 3 is missing, a retransmission request is initiated, and if the retransmission times out for 500 milliseconds, the permission migration instruction is triggered to reassign block 3 to other nodes; when the dynamic task state machine module reaches 50% of the task progress, the reassembly threshold of the low-risk task is automatically maintained as 2 / 3 according to the network delay fluctuation rate 15%; after the task is completed, the module releases the atomized permission unit and cleans up the blocks, without the need for redundant storage recovery; the whole process significantly reduces the system resource overhead by reducing the number of blocks, simplifying the conflict processing logic, and using a single-copy storage strategy, while ensuring the security and real-time requirements of low-risk tasks.

[0035] Although embodiments of the present application have been shown and described, it is to be understood that various modifications, substitutions, replacements and changes can be made to these embodiments without departing from the principles and spirit of the present application, and the scope of the present application is defined by the appended claims and their equivalents.

Claims

1. An edge computing-based multi-module collaborative management vocational education system, characterized in that, Comprise: A dynamic task state machine module for real-time monitoring of multi-dimensional state parameters such as task progress, network delay, device load, security risk level, data integrity and time sensitivity, and triggering authority migration instructions based on the comprehensive evaluation results of the multi-dimensional state parameters; The load threshold of the device load includes: the first load threshold is set to 70% of the safe upper limit of the CPU utilization rate of the computing unit, and the edge collaboration mode is allowed to be triggered when the device load is lower than this value; the second load threshold is set to 90% of the maximum safe load value, and the resource protection state is forced to enter when the device load exceeds this value; the third load threshold is set to 85% of the CPU utilization rate of the computing unit, and the computing unit is switched to the routing unit when the load exceeds this value; the load threshold is calibrated by historical data training and dynamic adjustment mechanism in the system initialization stage, and is dynamically fine-tuned during operation according to node performance and task type; The authority fragmentation edge cache module splits the authority into atomic authority units (APU) and divides it into several data blocks, stores it in the computing unit, routing unit and storage unit in the edge node cluster module according to the block strategy corresponding to the preset security level, generates a hash authority fingerprint chain formed by concatenating the block hash values in storage order, and verifies the authority integrity, wherein the security level includes key authorities corresponding to high-risk tasks and non-key authorities corresponding to low-risk tasks; The conflict resolution protocol module dynamically generates priority weights according to the remaining time of the task, preoccupies the authority operation right based on the priority weights, and records the timestamp, user identification and authority unit state of the conflict operation through the operation log chain, and executes local rollback; The edge node cluster module includes computing units, routing units and storage units with dynamic switching roles, and adjusts the node roles and authority storage paths according to the instructions of the dynamic task state machine module.

2. The multi-module collaborative management-based vocational education system of claim 1, wherein: The dynamic task state machine module includes: when the network delay exceeds the preset delay threshold and the device load is lower than the first load threshold, the current task security level is high risk, then trigger the edge collaboration mode, migrate the APU from the high delay node to the low delay node; when the device load exceeds the second load threshold, trigger the resource protection state, freeze the non-key APU and limit the write operation to the APU; when the network delay exceeds the preset delay threshold, the device load is lower than the first load threshold, and the device load exceeds the second load threshold at the same time, execute according to the priority of the resource protection state higher than the edge collaboration mode; when the task progress reaches the preset progress threshold and the data integrity check of the authority fragmentation edge cache module passes, switch to the task completion state and release all APUs; the non-key APU authority includes data query and material upload authority, and the key APU authority includes fund approval and device control authority.

3. The multi-module collaborative management-based vocational education system of claim 2, wherein: The permission fragmentation edge cache module dynamically adjusts the number of blocks and the reorganization threshold according to the task security level defined by the dynamic task state machine module, the task security level includes high-risk tasks and low-risk tasks, wherein the high-risk task is an operation involving fund approval or equipment control, and the low-risk task is an operation involving data query or material upload; The number of blocks of the high-risk task is a first number of blocks corresponding to a key permission, and the reorganization threshold is a first reorganization threshold; The number of blocks of the low-risk task is a second number of blocks corresponding to a non-key permission, and the reorganization threshold is a second reorganization threshold; The hash permission fingerprint chain is generated by concatenating the hash values of the block data in the dynamic index order when the block data is stored, and the block-by-block verification is required when the permission is verified. If the reorganization threshold of the continuous block verification is met, the complete permission unit is reorganized, otherwise the permission migration instruction is triggered.

4. The multi-module collaborative management-based vocational education system of claim 3, wherein: In the permission fragmentation edge cache module, when a user or a task initiates an access request to the APU, the APU block data stored in the computing unit, the routing unit and the storage unit is collected by the nearest edge node, the local hash value of each block is calculated and compared with the hash value of the corresponding node in the hash permission fingerprint chain; If the blocks that continuously meet the reorganization threshold pass the verification, the complete APU is reorganized and the access permission is granted; If the verification fails, the permission migration instruction is triggered, and the APU block data is redistributed to other edge nodes; the nearest edge node is the edge node with the lowest network delay of the user terminal.

5. The multi-module collaborative management-based vocational education system of claim 1, wherein: The conflict resolution protocol module dynamically generates priority weights according to the task remaining time, the remaining time is the difference between the task deadline and the current time, and the shorter the remaining time, the higher the priority weight; in the preemption operation, the high-priority weight task forcibly interrupts the permission access of the low-priority weight task, and adds its operation request to the processing queue; The operation log chain records the timestamp, user identifier and permission unit state of the conflict operation, the permission unit state includes permission unit identifier, operation type and operation result; when rolling back, the task state is recovered to the last time when the data integrity check is passed in reverse according to the operation log chain, and after rolling back, the integrity check process of the permission fragmentation edge cache module needs to be executed again.

6. The multi-module collaborative management-based vocational education system of claim 5, wherein: The generation and rollback of the operation log chain include: each operation record contains a forward hash pointer generated by the current block hash value and the previous pointer value, and an encrypted timestamp generated by using an asymmetric encryption algorithm; when rolling back, the continuity of the forward hash pointer needs to be decrypted and verified in reverse order of the generation time of the operation record; if it is detected that the hash values of adjacent blocks do not match or the timestamp decryption fails, the rollback is terminated and a safety alarm is triggered.

7. The multi-module collaborative management-based vocational education system of claim 1, wherein: In the edge node cluster module, the computing unit turns into the routing unit to optimize the communication path when the device load exceeds a third load threshold, the routing unit resumes as the computing unit when the network delay is lower than the delay threshold, and the computing task is re-executed; the storage unit dynamically allocates a storage path according to the APU block data storage requirement, and triggers a backup node takeover when a node disconnection is detected.

8. The multi-module collaborative management-based vocational education system of claim 7, wherein: The communication path optimization method of the edge node cluster module includes: the routing unit calculates the shortest delay path based on the network topology, dynamically adjusts the transmission order of the APU block data; when detecting path congestion, the routing unit splits the APU block data into sub-flows according to the block sequence number, and transmits in parallel through a plurality of low-delay paths selected dynamically; the storage unit recombines the data according to the block sequence number, if a block loss or out-of-order is detected, a block retransmission request is triggered, if the block retransmission request is not completed within a preset time, a permission migration instruction is triggered; the block sequence number is a unique sequence number allocated to the block data according to the dynamic index order.

9. The multi-module collaborative management-based vocational education system of claim 1, wherein: The state jump of the dynamic task state machine module first executes the resource protection state, and then executes the edge coordination mode; The state jump of the dynamic task state machine module triggers the permission block migration instruction; the integrity check result of the permission fragmentation edge cache module is fed back to the conflict resolution protocol module to adjust the priority weight; The rollback operation of the conflict resolution protocol module synchronously updates the progress parameter of the dynamic task state machine module.