Digital asset updating system and method
The digital asset update system, which utilizes intelligent prediction and distributed processing, solves the problems of low update efficiency and poor compatibility, enabling rapid and automatic digital asset updates and efficient platform adaptation, thereby improving the system's stability and compatibility.
Patent Information
- Application Number
- CN202511085540.9
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-08-04
- Publication Date
- 2025-11-14
AI Technical Summary
Existing digital asset update systems suffer from low update efficiency and poor compatibility. In particular, server pressure is too high during large-scale digital asset updates, and automatic adaptation to different platforms and devices is difficult to achieve.
The system employs an intelligent predictive update module for big data analysis and artificial intelligence prediction. Combined with a distributed intelligent update architecture, update tasks are distributed across multiple nodes for processing. An adaptive format conversion and compatibility module automatically identifies and converts digital asset formats. An update effect verification and backtracking module verifies the integrity and accuracy of the update, records anomalies, and restores the system to a stable version.
It improves the processing speed of large-scale digital asset updates, reduces user waiting time, enhances compatibility with new digital asset formats and platforms, reduces human error and operating costs, and strengthens system stability and fault tolerance.
Smart Images
Figure CN120950256A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of digital asset technology, specifically to a digital asset update system and method. Background Technology
[0002] In today's digital age, digital assets have become a core element of development across various industries, widely used in fields such as gaming, finance, media, and e-commerce. Digital assets include, but are not limited to, virtual items, digital currencies, digital copyrights, and electronic documents, and their value continues to increase as their application scenarios expand.
[0003] With the rapid increase in the types and quantities of digital assets, and the growing demands from users for real-time performance, security, and availability, timely, efficient, and secure updates to digital assets have become crucial. A robust digital asset update system can ensure the consistency and effectiveness of digital assets across different platforms and scenarios, enhance user experience, and promote the healthy development of the digital economy. For example, in the gaming industry, timely updates to in-game virtual items and scenes can maintain player activity and loyalty; in the financial sector, accurate and rapid updates to relevant cryptocurrency data are key to ensuring secure and smooth transactions.
[0004] However, current digital asset update systems have many problems, which seriously affect the update efficiency and compatibility of digital assets, as follows:
[0005] Inefficient Updates: Some digital asset update systems use traditional centralized update methods. When a large number of digital assets need to be updated, the server load becomes too high, resulting in slow update speeds and delays. For example, when a large online game releases a new version, millions of players simultaneously need to update in-game virtual items and scenes. The traditional centralized update system struggles to handle the massive concurrent requests, causing many players to wait for hours or even longer to complete the update, severely impacting their gaming experience.
[0006] Poor compatibility: Different types of digital assets have varying formats, and different application platforms and devices have different requirements for digital assets. Current update systems struggle to achieve compatibility with multiple digital asset formats and platforms. For example, when updating a digital media company's digital copyright assets to different playback platforms, the existing update system cannot automatically adapt because each platform has different requirements for video formats, resolutions, etc. Staff need to manually convert and adjust the formats, which not only increases workload but also increases the risk of errors.
[0007] Based on the above, a system and method for updating digital assets are invented. Summary of the Invention
[0008] To address the aforementioned technical problems, according to one aspect of the present invention, the present invention provides the following technical solution:
[0009] A digital asset update system, comprising:
[0010] The intelligent prediction and update module is used to analyze and predict information about digital assets using big data analytics and artificial intelligence technologies, in order to anticipate the update needs of digital assets in advance.
[0011] The distributed intelligent update architecture module is used to distribute digital asset update tasks to multiple nodes for processing based on the prediction results of the intelligent prediction update module. This avoids the problem of excessive server pressure caused by centralized updates. At the same time, it dynamically allocates update tasks according to the load of the nodes to improve the overall update efficiency.
[0012] The adaptive format conversion and compatibility module is used to automatically identify the format of different types of digital assets based on the parsing and conversion engine of multiple digital asset formats before the distributed nodes execute specific update tasks. It can also automatically convert and adapt the format according to the requirements of the target platform and device. At the same time, it has self-learning capabilities to continuously optimize the format conversion algorithm by analyzing historical update data, thereby improving compatibility with new digital asset formats and new platforms.
[0013] The update effect verification and backtracking module is used to automatically compare the asset data characteristics before and after the data update is completed on the distributed nodes to verify the completeness and accuracy of the update. Then, when an update anomaly is found during verification, the version backtracking mechanism can be automatically triggered to quickly restore the stable version before the update and record the reason for the anomaly. After that, the complete update log will be stored based on blockchain technology.
[0014] In a preferred embodiment of the digital asset update system described in this invention, the intelligent predictive update module includes:
[0015] The multi-dimensional data acquisition module is used to collect user behavior data, asset attribute data, environmental correlation data, and data interface data.
[0016] The data preprocessing module is used to perform cleaning, noise reduction, standardization, and feature engineering preprocessing operations on the collected data.
[0017] The prediction model training and optimization module is used to first select the model, then perform training iterations, and finally perform dynamic calibration.
[0018] Update the demand forecast generation module to first calculate the demand probability, then perform segmented scenario forecasting, and finally visualize the forecast results.
[0019] The pre-update preparation module is used to preload resources, then perform conflict detection, and finally schedule tasks.
[0020] In a preferred embodiment of the digital asset update system described in this invention, the multi-dimensional data acquisition module includes:
[0021] The user behavior data module is used to collect user operation records of digital assets through tracking technology.
[0022] The asset attribute data module is used to collect the static and dynamic attributes of digital assets themselves;
[0023] The environment-related data module is used to synchronize external environment information;
[0024] The data interface module is used to obtain cross data through the interface data.
[0025] In a preferred embodiment of the digital asset update system described in this invention, the data preprocessing module includes:
[0026] The cleaning and noise reduction module is used to remove outliers, fill in missing values, and filter invalid data.
[0027] The standardization conversion module is used to first unify the format of data from different dimensions, and then eliminate the differences in units through normalization processing;
[0028] The feature engineering module is used to build derived features.
[0029] In a preferred embodiment of the digital asset update system described in this invention, the prediction model training and optimization module includes:
[0030] The model selection module is used to adopt a hybrid algorithm architecture. The base layer uses a time series model to predict periodic update requirements, the enhancement layer overlays a machine learning model to analyze nonlinear factors, and the real-time layer introduces an LSTM neural network to process high-dimensional user behavior data.
[0031] The training iteration module is used to perform rolling training with historical data as the training set and divide the time window into weeks to adjust the model parameters in real time through the validation set.
[0032] The dynamic calibration module is used to receive feedback data from the update effect verification and backtracking modules, and optimizes the model weights through a reinforcement learning mechanism.
[0033] In a preferred embodiment of the digital asset update system described in this invention, the update demand prediction and generation module includes:
[0034] The demand probability calculation module is used to enable the model to output the update probability of a single digital asset in the next 72 hours, and generate a priority ranking table by combining the asset importance weight;
[0035] The scenario-specific prediction module is used to output specific prediction results for different application scenarios;
[0036] The prediction results visualization module is used to generate dynamic heatmaps and timeline schedules, and push them to the distributed intelligent update architecture module through a data interface.
[0037] In a preferred embodiment of the digital asset update system described in this invention, the pre-update preparation module includes:
[0038] The resource preloading module is used to distribute update packages for high-priority assets to edge nodes in advance based on prediction results, thereby reducing download latency during subsequent updates.
[0039] The conflict detection module is used to predict possible update conflicts and trigger the pre-conversion process of the adaptive format conversion and compatibility module in advance to generate alternative format schemes.
[0040] The task scheduling module is used to output suggested execution periods for update tasks to the distributed intelligent update architecture module, and reserve resource redundancy to cope with unexpected sudden update needs.
[0041] As a preferred embodiment of the digital asset update system described in this invention, the distributed intelligent update architecture module includes:
[0042] The task receiving and parsing module is used to first receive the update task list output by the intelligent prediction update module through the API interface; then it automatically extracts the key parameters of the task, generates standardized task instructions, and associates them with the preloaded resource address of the asset; then it verifies the identity of the task source to ensure that the task instructions have not been tampered with, and at the same time parses the encrypted signature attached to the task to verify the legality of the task.
[0043] The node resource evaluation and screening module is used to first scan the resource data of all available nodes in the distributed node network in real time; then, it uses a weighted algorithm to give a comprehensive score to the nodes; and finally, it selects suitable nodes according to task priority.
[0044] The task sharding and dynamic allocation module is used to first split large digital assets into several sub-tasks using a hash sharding algorithm and generate a unique verification value for each shard; then, based on node scores and task priorities, sub-tasks are allocated through a dynamic scheduling algorithm; finally, a smart contract containing sub-task details, completion deadlines, data encryption keys, and reward and punishment rules is generated for each node, and the contract is automatically triggered to take effect after the node confirms receipt, ensuring the traceability of task execution.
[0045] The inter-node data synchronization and consistency verification module is used to first obtain the corresponding asset shard data from edge nodes or adjacent nodes after receiving the subtask. It uses a P2P protocol to accelerate the transmission and protects the transmission process with symmetric encryption. Then, after each node completes the sharding process, it writes the processing result into the local blockchain ledger and broadcasts it to all nodes in the network. It uses a Byzantine fault tolerance algorithm to achieve consensus. When more than 2 / 3 of the nodes confirm that the data is consistent, the shard update is deemed valid. Afterwards, when node data inconsistency occurs, a secondary verification is automatically triggered. The three nodes with the highest scores recalculate and arbitrate, and the majority result is used as the final value. At the same time, abnormal nodes are marked and their subsequent task allocation weight is reduced.
[0046] The task execution and real-time monitoring module is used to first enable each node to perform update operations on the allocated asset shards according to the smart contract, supporting multi-threaded parallel processing and recording operation logs for each step; then, the distributed monitoring system collects node data in real time and generates a visual dashboard; subsequently, when a node fails or a task times out, the backup node takeover mechanism is automatically triggered, extracting unfinished subtasks from the task contract and reassigning them to the backup node to ensure that the overall task progress is not affected.
[0047] The task settlement and resource release module is used to first summarize the update results of all shards after all nodes have completed their sub-tasks, generate the final hash value of the complete asset, and compare it with the target hash value preset by the intelligent prediction update module. If the verification passes, the update is considered successful. Then, based on the smart contract, automatic settlement is performed, and computing power points are awarded to nodes that complete the task on time and whose data is consistent. Points are deducted from abnormal nodes and their task participation rights are restricted for 24 hours. After the update task is completed, the computing power and storage resources of the nodes are automatically released, and the final update result is synchronized to the update effect verification and backtracking module. At the same time, the local temporary task cache is cleared to reserve resources for the next round of tasks.
[0048] A method for updating digital assets includes the following specific steps:
[0049] S1: The intelligent prediction update module uses big data analysis and artificial intelligence technology to analyze and predict information about digital assets in order to determine the update needs of digital assets in advance.
[0050] S2: Based on the prediction results of the intelligent prediction update module, the distributed intelligent update architecture module uses a distributed node network to distribute digital asset update tasks to multiple nodes for processing, avoiding the problem of excessive server pressure caused by centralized updates. At the same time, it will dynamically allocate update tasks according to the load of the nodes to improve the overall update efficiency.
[0051] S3: Before executing specific update tasks on distributed nodes, the adaptive format conversion and compatibility module can automatically identify the formats of different types of digital assets based on the parsing and conversion engine of multiple digital asset formats, and automatically perform format conversion and adaptation according to the requirements of the target platform and device. At the same time, it has self-learning capabilities to continuously optimize the format conversion algorithm by analyzing historical update data, thereby improving compatibility with new digital asset formats and new platforms.
[0052] S4: After the data update is completed on the distributed nodes through the update effect verification and backtracking module, the asset data characteristics before and after the update are automatically compared to verify the completeness and accuracy of the update. Then, when an update anomaly is found during the verification, the version backtracking mechanism can be automatically triggered to quickly restore the stable version before the update and record the reason for the anomaly. After that, the complete update log will be stored based on blockchain technology.
[0053] Compared with existing technologies:
[0054] I. Advantages in addressing the problem of "low update efficiency"
[0055] The distributed intelligent update architecture module effectively solves the problem of excessive server load in traditional centralized updates by distributing update tasks across multiple nodes. Data synchronization and consistency verification are achieved between nodes based on blockchain technology. Combined with a dynamic task allocation mechanism, the workload can be flexibly adjusted according to node load, significantly improving the processing speed for concurrent updates of large-scale digital assets and avoiding long waiting times similar to those experienced by millions of players updating simultaneously in large online games.
[0056] The intelligent predictive update module leverages big data analytics and artificial intelligence to anticipate and prepare for the update needs of digital assets in advance, reducing user wait times. For example, in a gaming scenario, it prepares updates for high-demand virtual items in advance, allowing for rapid updates when players require them, further optimizing update efficiency.
[0057] II. Advantages in addressing the "poor compatibility" issue
[0058] The adaptive format conversion and compatibility module utilizes multiple format parsing and conversion engines to automatically identify and convert digital asset formats. This eliminates the need for manual intervention, adapting to different platforms and devices and resolving the issue of manual format conversion required by digital media companies when updating digital copyright assets across multiple platforms. Furthermore, the module's self-learning capability continuously optimizes its algorithms by analyzing historical data, enhancing compatibility with new digital asset formats and platforms, reducing human error, and lowering operational costs. Attached Figure Description
[0059] Figure 1This is a schematic diagram of the overall framework of the present invention;
[0060] Figure 2 This is a schematic diagram of the intelligent prediction and update module framework of the present invention;
[0061] Figure 3 This is a schematic diagram of the distributed intelligent update architecture module framework of the present invention. Detailed Implementation
[0062] To make the objectives, technical solutions, and advantages of the present invention clearer, the embodiments of the present invention will be described in further detail below with reference to the accompanying drawings.
[0063] This invention provides a digital asset update system; please refer to [link / reference]. Figures 1-3 ,include:
[0064] The intelligent prediction and update module uses big data analytics and artificial intelligence to analyze and predict information such as the frequency of use of digital assets and changes in user demand, so as to determine the update needs of digital assets in advance. For example, based on the behavior data of game players, it can predict which virtual items will have high demand in the near future, prepare to update these items in advance, and quickly complete the update when users need it, reducing waiting time.
[0065] The distributed intelligent update architecture module is used to distribute digital asset update tasks to multiple nodes for processing based on the prediction results of the intelligent prediction update module. This avoids the problem of excessive server pressure caused by centralized updates. At the same time, it dynamically allocates update tasks according to the load of the nodes to improve the overall update efficiency.
[0066] The adaptive format conversion and compatibility module is used to automatically identify the format of different types of digital assets based on the parsing and conversion engine of multiple digital asset formats before the distributed nodes execute specific update tasks. It can also automatically convert and adapt the format according to the requirements of the target platform and device. At the same time, it has self-learning capabilities to continuously optimize the format conversion algorithm by analyzing historical update data, thereby improving compatibility with new digital asset formats and new platforms.
[0067] The update effect verification and backtracking module is used to automatically compare the asset data characteristics (such as hash value, metadata identifier, functional parameters, etc.) before and after the data update is completed on the distributed nodes to verify the completeness and accuracy of the update. Then, when the verification finds an update anomaly (such as missing data or functional failure), it can automatically trigger the version backtracking mechanism to quickly restore to the stable version before the update and record the reason for the anomaly. Afterwards, it will store the complete update log based on blockchain technology, including update node, time, data change details, etc., to support full-link traceability analysis of abnormal updates.
[0068] The advantages of updating the effect verification and backtracking module are:
[0069] 1. Address the potential problem of "difficulty in quickly correcting anomalies after updates" in existing solutions: For example, if a balance calculation error occurs after a digital currency update on a financial platform, this module can complete the verification and backtrack to the correct version within 5 minutes to prevent further loss of user assets.
[0070] 2. Enhance system fault tolerance: By performing real-time verification and rapid backtracking, the risk of update failure caused by distributed node synchronization deviations, format conversion errors, etc., is reduced, thereby improving system stability.
[0071] The intelligent prediction and update module includes:
[0072] The multi-dimensional data acquisition module is used to collect user behavior data, asset attribute data, environmental correlation data, and data interface data.
[0073] The data preprocessing module is used to perform cleaning, noise reduction, standardization, and feature engineering preprocessing operations on the collected data.
[0074] The prediction model training and optimization module is used to first select the model, then perform training iterations, and finally perform dynamic calibration.
[0075] Update the demand forecast generation module to first calculate the demand probability, then perform segmented scenario forecasting, and finally visualize the forecast results.
[0076] The pre-update preparation module is used to preload resources, then perform conflict detection, and finally schedule tasks.
[0077] The multi-dimensional data acquisition module includes:
[0078] The user behavior data module is used to collect user operation records of digital assets through tracking technology, including access frequency, usage duration, interaction depth (such as the number of times game items are equipped, the frequency of digital currency transactions), dwell time, etc., covering data from multiple terminals such as PC, mobile, and blockchain wallet.
[0079] The asset attribute data module is used to collect the static attributes (such as format type, size, and creation time) and dynamic attributes (such as version iteration records, historical update frequency, and current validity period) of digital assets.
[0080] The environmental data module is used to synchronize external environmental information, including the target platform's version update plan (such as game engine upgrade time), the geographical distribution and time zone characteristics of the user group, and industry hot events (such as the impact of changes in financial policies on digital currencies).
[0081] The data interface module is used to obtain cross-data such as the execution efficiency of historical update tasks and update request keywords in user feedback through interface data.
[0082] The data preprocessing module includes:
[0083] The cleaning and noise reduction module is used to remove outliers (such as instantaneous high-frequency access data caused by accidental operation), fill in missing values (using time series interpolation to process discontinuous user behavior records), and filter invalid data (such as duplicate asset attribute entries).
[0084] The standardization conversion module is used to first unify the format of data from different dimensions (such as converting user access time into timestamps and unifying asset size into MB units), and then eliminate the differences in units through normalization (such as mapping usage frequency to the 0-1 range).
[0085] The feature engineering module is used to build derived features, such as calculating the commercial value weight of an asset by "usage frequency × user payment level" and estimating the potential update cycle by "historical update interval × current version lifecycle".
[0086] The prediction model training and optimization module includes:
[0087] The model selection module is used to adopt a hybrid algorithm architecture. The base layer uses time series models (such as ARIMA) to predict periodic update needs (such as monthly digital copyright renewals). The enhancement layer overlays machine learning models (such as random forests) to analyze nonlinear factors (such as the surge in demand for virtual items due to sudden hot events). The real-time layer introduces LSTM neural networks to process high-dimensional user behavior data (such as minute-level fluctuations in game item interactions).
[0088] The training iteration module is used to train on historical data and divide the time window into weeks for rolling training, so as to adjust the model parameters (such as the number of hidden layer nodes of LSTM and the tree depth of random forest) in real time through the validation set (accounting for 20%).
[0089] The dynamic calibration module is used to access feedback data from the update effect verification and backtracking module (such as prediction accuracy and actual user satisfaction after the update), and optimizes the model weights through reinforcement learning mechanisms (such as increasing the weight of corresponding features for scenarios with prediction deviation rates exceeding 15%).
[0090] The update requirement prediction and generation module includes:
[0091] The demand probability calculation module is used to enable the model to output the update probability of a single digital asset in the next 72 hours (e.g., the update probability of a certain virtual skin is 89%), and generate a priority ranking table by combining the asset importance weight (e.g., the weight of core financial data is higher than that of ordinary image materials);
[0092] The segmented scenario prediction module is used to output specific prediction results for different application scenarios. For example, in the game scenario, items marked as "commonly used by users with high churn risk" need to be updated first, and in the financial scenario, "digital currency contracts that are about to expire" need to be prepared for updates in advance.
[0093] The prediction results visualization module is used to generate dynamic heatmaps (showing the intensity of update demand for different assets) and timeline schedules (marking recommended update periods), and push them to the distributed intelligent update architecture module through the data interface.
[0094] The pre-update preparation module includes:
[0095] The resource preloading module is used to distribute update packages of high-priority assets to edge nodes (such as distributed servers in the user's region) in advance based on the prediction results, thereby reducing download latency during subsequent updates.
[0096] The conflict detection module is used to predict possible update conflicts (such as the potential compatibility issues between a digital asset and a new version of the target platform), and to trigger the pre-conversion process of the adaptive format conversion and compatibility module in advance to generate alternative format solutions.
[0097] The task scheduling module is used to output suggested execution times for update tasks to the distributed intelligent update architecture module (such as avoiding peak user access times from 2-4 am), and to reserve resource redundancy (such as bandwidth and computing power) to cope with unexpected sudden update needs.
[0098] The distributed intelligent update architecture module includes:
[0099] The task receiving and parsing module first receives the update task list output by the intelligent prediction update module via the API interface, including information such as the ID of the digital asset to be updated, priority, suggested execution time period, and target terminal type (e.g., PC / mobile / blockchain node). Then, it automatically extracts key task parameters, generates standardized task instructions (e.g., "Asset ID: V202306_001, priority: high, full user update must be completed within 2 hours"), and associates it with the preloaded resource address of the asset (edge node storage path). After that, it verifies the identity of the task source (based on blockchain node certificate verification) to ensure that the task instructions have not been tampered with, and parses the encrypted signature attached to the task (using an asymmetric encryption algorithm) to verify the legality of the task.
[0100] The node resource assessment and screening module first scans the resource data of all available nodes in the distributed node network in real time, including current computing power load (CPU / memory utilization), network bandwidth (uplink / downlink speed), storage capacity, historical task success rate, and geographical location (physical distance from the target user group). Then, it uses a weighted algorithm to give nodes a comprehensive score (out of 100), with computing power load (30%), network bandwidth (25%), historical success rate (20%), physical distance (15%), and storage capacity (10%) as the core indicators. After that, it selects suitable nodes according to task priority. High-priority tasks (such as core financial data updates) are matched with nodes with a score ≥80, while ordinary tasks (such as image material updates) can be relaxed to ≥60. At the same time, it ensures that each task is allocated at least 3 backup nodes to cope with sudden failures.
[0101] The task sharding and dynamic allocation module is used to first split large digital assets (such as multi-GB game installation packages) into several sub-tasks (single fragment size ≤ 100MB) using a hash sharding algorithm, and generate a unique checksum (SHA-256 hash) for each fragment. Then, based on node scores and task priorities, the sub-tasks are allocated using a dynamic scheduling algorithm (such as an improved Round-Robin algorithm). For example, core fragments of high-priority assets are allocated to low-load nodes, while limiting the maximum number of tasks on a single node to no more than 60% of its computing power to avoid resource overload. Afterward, a smart contract (based on blockchain smart contract technology) is generated for each node, containing sub-task details, completion deadlines, data encryption keys, and reward and punishment rules. The contract is automatically triggered after the node confirms receipt, ensuring the traceability of task execution.
[0102] The inter-node data synchronization and consistency verification module is used to first obtain the corresponding asset shard data from edge nodes or adjacent nodes after a node receives a subtask. It uses a P2P protocol to accelerate transmission (such as BitTorrent technology) and protects the transmission process with symmetric encryption (AES-256). Then, after each node completes the sharding process, it writes the processing result (such as the updated asset hash value and execution timestamp) into the local blockchain ledger and broadcasts it to all nodes in the network. A Byzantine fault-tolerant algorithm is used to achieve consensus; when more than 2 / 3 of the nodes confirm data consistency, the shard update is deemed valid. Subsequently, if node data inconsistency occurs (such as hash mismatch), a secondary verification is automatically triggered. The three nodes with the highest scores recalculate and arbitrate, using the majority result as the final value. At the same time, abnormal nodes are marked and their subsequent task allocation weight is reduced.
[0103] The task execution and real-time monitoring module is used to first enable each node to perform update operations (such as replacing the old version or writing new attributes) on the allocated asset shards according to the smart contract. It supports multi-threaded parallel processing and records the operation logs for each step (including intermediate states, time consumption, error codes, etc.). Then, a distributed monitoring system (such as one built on Prometheus + Grafana) collects data on the task progress of the nodes in real time (such as "8 / 10 shards completed"), resource usage changes, network latency, etc., and generates a visual dashboard. Afterwards, when a node fails (such as offline, sudden drop in computing power) or the task times out (more than 15 minutes beyond the agreed time), the backup node takeover mechanism is automatically triggered. The unfinished sub-tasks are extracted from the task contract and reassigned to the backup node to ensure that the overall task progress is not affected.
[0104] The task settlement and resource release module is used to first summarize the update results of all shards after all nodes have completed their sub-tasks, generate the final hash value of the complete asset, and compare it with the target hash value preset by the intelligent prediction update module. If the verification passes, the update is considered successful. Then, based on the smart contract, automatic settlement is performed. Nodes that complete the task on time and whose data is consistent are awarded computing power points (which can be used to increase the priority of subsequent tasks). Nodes with abnormal data have points deducted and their task participation rights are restricted for 24 hours. After the update task is completed, the computing power and storage resources of the nodes are automatically released (temporary shard files are deleted), and the final update result is synchronized to the update effect verification and backtracking module. At the same time, the local temporary task cache is cleared to reserve resources for the next round of tasks.
[0105] A method for updating digital assets includes the following specific steps:
[0106] S1: The intelligent prediction update module uses big data analysis and artificial intelligence technology to analyze and predict information about digital assets in order to determine the update needs of digital assets in advance.
[0107] S2: Based on the prediction results of the intelligent prediction update module, the distributed intelligent update architecture module uses a distributed node network to distribute digital asset update tasks to multiple nodes for processing, avoiding the problem of excessive server pressure caused by centralized updates. At the same time, it will dynamically allocate update tasks according to the load of the nodes to improve the overall update efficiency.
[0108] S3: Before executing specific update tasks on distributed nodes, the adaptive format conversion and compatibility module can automatically identify the formats of different types of digital assets based on the parsing and conversion engine of multiple digital asset formats, and automatically perform format conversion and adaptation according to the requirements of the target platform and device. At the same time, it has self-learning capabilities to continuously optimize the format conversion algorithm by analyzing historical update data, thereby improving compatibility with new digital asset formats and new platforms.
[0109] S4: After the data update is completed on the distributed nodes through the update effect verification and backtracking module, the asset data characteristics before and after the update are automatically compared to verify the completeness and accuracy of the update. Then, when an update anomaly is found during the verification, the version backtracking mechanism can be automatically triggered to quickly restore the stable version before the update and record the reason for the anomaly. After that, the complete update log will be stored based on blockchain technology.
[0110] Although the present invention has been described above with reference to embodiments, various modifications can be made and components can be replaced with equivalents without departing from the scope of the invention. In particular, as long as there is no structural conflict, the features in the disclosed embodiments can be combined with each other in any manner. The lack of an exhaustive description of these combinations in this specification is merely for the sake of brevity and resource conservation. Therefore, the present invention is not limited to the specific embodiments disclosed herein, but includes all technical solutions falling within the scope of the claims.
Claims
1. A digital asset update system, characterized in that, include: The intelligent prediction and update module is used to analyze and predict information about digital assets using big data analytics and artificial intelligence technologies, in order to anticipate the update needs of digital assets in advance. The distributed intelligent update architecture module is used to distribute digital asset update tasks to multiple nodes for processing based on the prediction results of the intelligent prediction update module. This avoids the problem of excessive server pressure caused by centralized updates. At the same time, it dynamically allocates update tasks according to the load of the nodes to improve the overall update efficiency. The adaptive format conversion and compatibility module is used to automatically identify the format of different types of digital assets based on the parsing and conversion engine of multiple digital asset formats before the distributed nodes execute specific update tasks. It can also automatically convert and adapt the format according to the requirements of the target platform and device. At the same time, it has self-learning capabilities to continuously optimize the format conversion algorithm by analyzing historical update data, thereby improving compatibility with new digital asset formats and new platforms. The update effect verification and backtracking module is used to automatically compare the asset data characteristics before and after the data update is completed on the distributed nodes to verify the completeness and accuracy of the update. Then, when an update anomaly is found during verification, the version backtracking mechanism can be automatically triggered to quickly restore the stable version before the update and record the reason for the anomaly. After that, the complete update log will be stored based on blockchain technology.
2. The digital asset update system according to claim 1, characterized in that, The intelligent prediction and update module includes: The multi-dimensional data acquisition module is used to collect user behavior data, asset attribute data, environmental correlation data, and data interface data. The data preprocessing module is used to perform cleaning, noise reduction, standardization, and feature engineering preprocessing operations on the collected data. The prediction model training and optimization module is used to first select the model, then perform training iterations, and finally perform dynamic calibration. Update the demand forecast generation module to first calculate the demand probability, then perform segmented scenario forecasting, and finally visualize the forecast results. The pre-update preparation module is used to preload resources, then perform conflict detection, and finally schedule tasks.
3. The digital asset update system according to claim 2, characterized in that, The multi-dimensional data acquisition module includes: The user behavior data module is used to collect user operation records of digital assets through tracking technology. The asset attribute data module is used to collect the static and dynamic attributes of digital assets themselves; The environment-related data module is used to synchronize external environment information; The data interface module is used to obtain cross data through the interface data.
4. A digital asset update system according to claim 2, characterized in that, The data preprocessing module includes: The cleaning and noise reduction module is used to remove outliers, fill in missing values, and filter invalid data. The standardization conversion module is used to first unify the format of data from different dimensions, and then eliminate the differences in units through normalization processing; The feature engineering module is used to build derived features.
5. A digital asset update system according to claim 2, characterized in that, The prediction model training and optimization module includes: The model selection module is used to adopt a hybrid algorithm architecture. The base layer uses a time series model to predict periodic update requirements, the enhancement layer overlays a machine learning model to analyze nonlinear factors, and the real-time layer introduces an LSTM neural network to process high-dimensional user behavior data. The training iteration module is used to perform rolling training with historical data as the training set and divide the time window into weeks to adjust the model parameters in real time through the validation set. The dynamic calibration module is used to receive feedback data from the update effect verification and backtracking modules, and optimizes the model weights through a reinforcement learning mechanism.
6. A digital asset update system according to claim 2, characterized in that, The update requirement prediction and generation module includes: The demand probability calculation module is used to enable the model to output the update probability of a single digital asset in the next 72 hours, and generate a priority ranking table by combining the asset importance weight; The scenario-specific prediction module is used to output specific prediction results for different application scenarios; The prediction results visualization module is used to generate dynamic heatmaps and timeline schedules, and push them to the distributed intelligent update architecture module through a data interface.
7. A digital asset update system according to claim 2, characterized in that, The pre-update preparation module includes: The resource preloading module is used to distribute update packages for high-priority assets to edge nodes in advance based on prediction results, thereby reducing download latency during subsequent updates. The conflict detection module is used to predict possible update conflicts and trigger the pre-conversion process of the adaptive format conversion and compatibility module in advance to generate alternative format schemes. The task scheduling module is used to output suggested execution periods for update tasks to the distributed intelligent update architecture module, and reserve resource redundancy to cope with unexpected sudden update needs.
8. A digital asset update system according to claim 1, characterized in that, The distributed intelligent update architecture module includes: The task receiving and parsing module is used to first receive the update task list output by the intelligent prediction update module through the API interface; then it automatically extracts the key parameters of the task, generates standardized task instructions, and associates them with the preloaded resource address of the asset; then it verifies the identity of the task source to ensure that the task instructions have not been tampered with, and at the same time parses the encrypted signature attached to the task to verify the legality of the task. The node resource evaluation and screening module is used to first scan the resource data of all available nodes in the distributed node network in real time; then, it uses a weighted algorithm to give a comprehensive score to the nodes; and finally, it selects suitable nodes according to task priority. The task sharding and dynamic allocation module is used to first split large digital assets into several sub-tasks using a hash sharding algorithm and generate a unique verification value for each shard; then, based on node scores and task priorities, sub-tasks are allocated through a dynamic scheduling algorithm; finally, a smart contract containing sub-task details, completion deadlines, data encryption keys, and reward and punishment rules is generated for each node, and the contract is automatically triggered to take effect after the node confirms receipt, ensuring the traceability of task execution. The inter-node data synchronization and consistency verification module is used to first obtain the corresponding asset shard data from edge nodes or adjacent nodes after receiving the subtask. It uses a P2P protocol to accelerate the transmission and protects the transmission process with symmetric encryption. Then, after each node completes the sharding process, it writes the processing result into the local blockchain ledger and broadcasts it to all nodes in the network. It uses a Byzantine fault tolerance algorithm to achieve consensus. When more than 2 / 3 of the nodes confirm that the data is consistent, the shard update is deemed valid. Afterwards, when node data inconsistency occurs, a secondary verification is automatically triggered. The three nodes with the highest scores recalculate and arbitrate, and the majority result is used as the final value. At the same time, abnormal nodes are marked and their subsequent task allocation weight is reduced. The task execution and real-time monitoring module is used to first enable each node to perform update operations on the allocated asset shards according to the smart contract, supporting multi-threaded parallel processing and recording operation logs for each step; then, the distributed monitoring system collects node data in real time and generates a visual dashboard; subsequently, when a node fails or a task times out, the backup node takeover mechanism is automatically triggered, extracting unfinished subtasks from the task contract and reassigning them to the backup node to ensure that the overall task progress is not affected. The task settlement and resource release module is used to first summarize the update results of all shards after all nodes have completed their sub-tasks, generate the final hash value of the complete asset, and compare it with the target hash value preset by the intelligent prediction update module. If the verification passes, the update is considered successful. Then, based on the smart contract, automatic settlement is performed, and computing power points are awarded to nodes that complete the task on time and whose data is consistent. Points are deducted from abnormal nodes and their task participation rights are restricted for 24 hours. After the update task is completed, the computing power and storage resources of the nodes are automatically released, and the final update result is synchronized to the update effect verification and backtracking module. At the same time, the local temporary task cache is cleared to reserve resources for the next round of tasks.
9. A method for updating digital assets, characterized in that, The specific steps are as follows: S1: The intelligent prediction update module uses big data analysis and artificial intelligence technology to analyze and predict information about digital assets in order to determine the update needs of digital assets in advance. S2: Based on the prediction results of the intelligent prediction update module, the distributed intelligent update architecture module uses a distributed node network to distribute digital asset update tasks to multiple nodes for processing, avoiding the problem of excessive server pressure caused by centralized updates. At the same time, it will dynamically allocate update tasks according to the load of the nodes to improve the overall update efficiency. S3: Before executing specific update tasks on distributed nodes, the adaptive format conversion and compatibility module can automatically identify the formats of different types of digital assets based on the parsing and conversion engine of multiple digital asset formats, and automatically perform format conversion and adaptation according to the requirements of the target platform and device. At the same time, it has self-learning capabilities to continuously optimize the format conversion algorithm by analyzing historical update data, thereby improving compatibility with new digital asset formats and new platforms. S4: After the data update is completed on the distributed nodes through the update effect verification and backtracking module, the asset data characteristics before and after the update are automatically compared to verify the completeness and accuracy of the update. Then, when an update anomaly is found during the verification, the version backtracking mechanism can be automatically triggered to quickly restore the stable version before the update and record the reason for the anomaly. After that, the complete update log will be stored based on blockchain technology.