Resource pool coordination system, method and device based on block chain and big data AI

Abstract

The invention discloses a resource pool coordination system, method and device based on a block chain and big data AI, and relates to the technical field of resource pool coordination, and the method comprises the following steps: before the resource pool coordination starts, establishing a unified time sequence monitoring baseline, carrying out the layering marking of an intelligent contract triggering link used for resource distribution in the block chain, and carrying out the monitoring of the unified time sequence monitoring baseline; acquiring depth parameters and time delay parameters called by each level in real time, and generating a recursive risk map containing dynamic constraint conditions based on an acquisition result; under the constraint of the recursion risk map, a force field calculation mechanism is introduced, and a micro-scale disturbance signal is injected into a smart contract trigger link to calculate a stability factor of a call chain. According to the method, dynamic monitoring and early deadlock recognition of the smart contract call chain are realized by constructing the unified time sequence monitoring baseline and the recursive risk map, and causal tracking and dynamic repair of the abnormal link are realized by combining the anti-fact playback chain and the self-adaptive decoupling mapping, so that the coordination efficiency of the block chain resource pool and the system stability are improved.

Description

Technical Field

[0001] This invention relates to the field of resource pool coordination technology, and more specifically to resource pool coordination systems, methods and apparatus based on blockchain and big data AI. Background Technology

[0002] Resource pool coordination based on blockchain and big data AI refers to using the decentralized, immutable, and traceable characteristics of blockchain to build a trusted ledger for resource sharing and scheduling in a multi-source, heterogeneous computing, storage, and network resource environment. Simultaneously, it combines the real-time analysis and predictive capabilities of big data AI to dynamically model and intelligently optimize the supply and demand status, utilization efficiency, and abnormal fluctuations within the resource pool. Its core significance lies in: blockchain ensuring the transparency and fairness of the resource scheduling process, preventing single points of failure and malicious manipulation; and big data AI achieving optimal resource allocation decisions through pattern recognition, trend prediction, and adaptive learning, thereby forming a self-organizing, adaptive, and self-evolving resource coordination mechanism in complex business scenarios. This approach not only significantly improves resource utilization and response speed but also builds a trusted sharing foundation in cross-organizational and cross-regional collaborative environments.

[0003] The existing technology has the following shortcomings: Blockchain-based resource pool coordination often relies on smart contracts to automatically trigger and execute resource allocation. However, in extremely sparse time periods, when resource demands trigger specific contract logic, the on-chain call chain may be abnormally lengthened, leading to a continuous accumulation of recursive calls and forming a hidden deadlock. Because this type of deadlock is not easily identified directly in the short term, its anomaly can be gradually masked over long periods of operation, freezing critical resource allocation channels and preventing some resources from being properly scheduled and utilized. Ultimately, the entire resource pool will experience local resource silos in a dynamically changing environment, not only disrupting the overall balance of resource sharing but also potentially triggering the spread of systemic risks.

[0004] The information disclosed in the background section is only intended to enhance the understanding of the background of this disclosure, and therefore may include information that does not constitute prior art known to those skilled in the art. Summary of the Invention

[0005] The purpose of this invention is to provide a resource pool coordination system, method, and apparatus based on blockchain and big data AI to solve the problems mentioned in the background art.

[0006] To achieve the above objectives, the present invention provides the following technical solution: a resource pool coordination method based on blockchain and big data AI, comprising the following steps: Step 1: Before resource pool coordination begins, establish a unified time-series monitoring baseline, hierarchically mark the smart contract triggering links used for resource allocation in the blockchain, collect the depth parameters and latency parameters of each level of call in real time, and generate a recursive risk map containing dynamic constraints based on the collection results to provide a quantitative reference for subsequent anomaly identification. Step 2: Under the constraints of the recursive risk graph, a force field calculation mechanism is introduced to inject micro-scale perturbation signals into the smart contract triggering link to calculate the stability factor of the calling chain. Based on the calculation results, local unstable segments are extracted and solidified to form potential deadlock fingerprints, so as to achieve the initial identification of recursive risks. Step 3: Based on the deadlock fingerprint, run the counterfactual replay chain to replay the historical contract triggering trajectory in the blockchain in a time sequence, strip away the normal call path and amplify the abnormal superposition link to generate an abnormal recursion chain model with causal traceability, so as to determine the formation mechanism of abnormal recursion and intervention strategy. Step 4: Based on the abnormal recursive chain model, construct an adaptive decoupling mapping, unlock the call chain that deviates from the normal state step by step and redirect it to a safe path. At the same time, write the correction results back to the unified timing monitoring baseline, thereby realizing the dynamic repair and global balance recovery of the resource allocation channel.

[0007] Preferably, the steps for establishing a unified time-series monitoring baseline include: Before resource pool coordination begins, the timestamp function of the blockchain ledger is called and combined with the external trusted time source interface to perform self-calibration and linear regression correction on the time offset of multiple nodes, and establish a unified time series monitoring baseline for global time alignment. After establishing a unified time-series monitoring baseline, the triggering links of smart contracts used for resource allocation in the blockchain are hierarchically marked, and hierarchical identifiers are generated by combining hash index and dependency matrix, so that they are associated with and stored with the timestamp of the unified time-series monitoring baseline. Based on the synchronization time data of the unified time series monitoring baseline, the call depth parameters and latency parameters of the smart contract triggering chain at each level are collected in real time, and a two-dimensional dynamic matrix of call depth-latency is constructed. Based on the dynamic changes of the matrix, a time-series clustering algorithm is used to calculate the dynamic constraint weights of risk clusters, generating a recursive risk map with a unified time-series monitoring baseline as the time dimension.

[0008] Preferably, the step of generating a recursive risk map includes: when calculating the dynamic constraint weights of risk clusters, using a time-series clustering algorithm with dynamically adjusted cluster centers to aggregate synchronously fluctuating links based on the call depth parameters and latency gradients of the trigger links of smart contracts at each level, to obtain time-series weight parameters that can reflect changes in the intensity of recursive risk, and using these weight parameters as the initial input values ​​for the subsequent force field calculation mechanism.

[0009] Preferably, the steps of introducing a force field calculation mechanism under the constraint of a recursive risk graph include: Based on the dynamic constraint weights and hierarchical structure in the recursive risk graph, each smart contract triggering link is regarded as a force-bearing unit, and a state evolution equation is established under a unified time-series monitoring baseline. After the force field calculation mechanism is initialized, microscale disturbance signals are injected into high-risk link areas and data on delay offset, call depth fluctuation and resource contention intensity changes are collected in real time. The stability factor of the link is calculated based on the collected data, and a dynamic response curve is generated. Locally unstable regions where the stability factor exceeds a threshold are extracted, and potential deadlock fingerprints are generated based on feature vector similarity to achieve initial identification of recursion risks.

[0010] Preferably, in the step of injecting microscale perturbation signals, the perturbation signals are non-destructive dynamic perturbation signals, and their injection frequency is synchronized with the block generation time to ensure that periodic perturbations are applied to the link's latency parameters, resource contention parameters, and state transition rate parameters without affecting the execution results of the smart contract logic.

[0011] Preferably, the steps for running a counterfactual replay chain based on deadlock fingerprints include: Based on the potential deadlock fingerprint, relevant historical block sequences are selected under a unified time-series monitoring baseline, and the input parameters, execution status, and intermediate state transition process of each smart contract call are replayed in time sequence. The replay results are analyzed using call depth and latency parameters. The dynamic constraint weights of the recursive risk graph are used to separate normal call paths and mark suspected abnormal links. Based on the execution time inversion and event dependency reconstruction of suspected abnormal links, a time-causality matrix is ​​constructed and a time evolution diagram of abnormal superimposed links is generated; Based on the time evolution diagram, an abnormal recursive chain model is established. Key nodes are extracted by calculating the transmission probability and influence intensity of event nodes, and corresponding intervention strategies are solidified.

[0012] Preferably, the step of establishing an abnormal recursive chain model includes: after generating the time evolution diagram of the abnormal superimposed link, calculating the transmission probability and influence intensity of each abnormal node based on the time-causality matrix, and comparing the propagation results under different intervention variables by introducing counterfactual simulation paths to determine the key node that can effectively block the abnormal recursive propagation, and solidifying the control conditions and intervention strategies of the key node into the abnormal recursive chain model.

[0013] Preferably, the steps for constructing an adaptive decoupling map based on the abnormal recursion chain model include: Based on the anomaly propagation path and key impact nodes in the anomaly recursive chain model, an anomaly propagation matrix is ​​established by combining the time series information of the unified time series monitoring baseline, and a link state phase space mapping is constructed to distinguish between anomaly states and steady states. The deviated link is unlocked step by step according to the degree of node deviation, and the priority unlocking target is determined by state similarity judgment and propagation probability calculation; For unlocked links, perform path redirection, guide them to a safe path using a dynamic path selection algorithm, and record path replacement events; The correction results are written back to the unified time-series monitoring baseline and the recursive risk map is updated synchronously to achieve dynamic repair and global balance restoration of resource allocation channels.

[0014] The resource pool coordination system based on blockchain and big data AI includes a time-series monitoring baseline construction module, a force field calculation and analysis module, a counterfactual replay chain modeling module, and an adaptive decoupling and repair module. The timing monitoring baseline construction module establishes a unified timing monitoring baseline before resource pool coordination begins. It hierarchically marks the smart contract triggering links used for resource allocation in the blockchain, collects the depth parameters and latency parameters of each level of call in real time, and generates a recursive risk map containing dynamic constraints based on the collection results. The force field calculation and analysis module, under the constraint of the recursive risk graph, introduces a force field calculation mechanism to inject micro-scale perturbation signals into the smart contract triggering link to calculate the stability factor of the calling chain, and extracts local unstable sections based on the calculation results, and solidifies the sections to form potential deadlock fingerprints. The counterfactual replay chain modeling module runs a counterfactual replay chain based on deadlock fingerprints, replays the historical contract triggering trajectory in the blockchain in a time sequence, strips out the normal call path and amplifies the abnormal superimposed link, and generates an abnormal recursive chain model with causal traceability. The adaptive decoupling repair module constructs an adaptive decoupling mapping based on the abnormal recursive chain model, unlocks and redirects call chains that deviate from the normal state to safe paths step by step, and writes the correction results back to the unified timing monitoring baseline.

[0015] The technical effects and advantages provided by the present invention in the above technical solution are as follows: This invention establishes a unified time-series monitoring baseline before resource pool coordination and combines a recursive risk graph and force field calculation mechanism to achieve real-time quantitative modeling and dynamic stability assessment of smart contract call chains during blockchain resource allocation. Through hierarchical labeling, time-series data acquisition, and micro-scale perturbation calculation, the latency fluctuations and recursion depth of each call chain can be continuously monitored, enabling early identification and location of potential deadlocks before anomalies spread. This mechanism not only effectively avoids resource locking and channel freezing caused by long-term call chain stacking but also significantly improves the scheduling transparency and response agility of the resource pool in multi-source heterogeneous environments, transforming the resource coordination process from static monitoring to an intelligent mode of proactive defense and adaptive prediction.

[0016] This invention achieves causal tracing, path repair, and global equilibrium restoration of abnormal recursive chains by introducing a counterfactual replay chain and adaptive decoupling mapping after anomaly identification. Through temporal replay of historical contract trajectories and amplified analysis of abnormal links, combined with adaptive unlocking and secure path redirection mechanisms, recursive deadlocks can be dynamically resolved and resource liquidity restored without interfering with the normal operation of the blockchain. The repair results are synchronously written back to a unified temporal monitoring baseline, enabling the system to have continuous self-correction capabilities, forming a closed-loop anomaly detection-intervention-repair mechanism. This technology not only improves the continuity of resource allocation and the robustness of system operation but also constructs a cross-level, traceable, and verifiable resource coordination system, effectively enhancing the global stability and long-term operational reliability of the blockchain resource pool. Attached Figure Description

[0017] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in this invention. For those skilled in the art, other drawings can be obtained based on these drawings.

[0018] Figure 1 This is a flowchart of the resource pool coordination method based on blockchain and big data AI of the present invention.

[0019] Figure 2 This is a schematic diagram of the module of the resource pool coordination system based on blockchain and big data AI of the present invention. Detailed Implementation

[0020] Exemplary embodiments will now be described more fully with reference to the accompanying drawings. However, these exemplary embodiments can be implemented in many forms and should not be construed as limited to the examples set forth herein; rather, they are provided so that the description of this disclosure will be more complete and fully convey the concept of the exemplary embodiments to those skilled in the art.

[0021] This invention provides, for example Figure 1 The resource pool coordination method based on blockchain and big data AI, as shown, includes the following steps: Step 1: Before resource pool coordination begins, establish a unified time-series monitoring baseline, hierarchically mark the smart contract triggering links used for resource allocation in the blockchain, collect the depth parameters and latency parameters of each level of call in real time, and generate a recursive risk map containing dynamic constraints based on the collection results to provide a quantitative reference for subsequent anomaly identification. The specific implementation steps are as follows: Before resource pool coordination begins, a unified time-series monitoring baseline is established for the allocation process of computing, storage, and network resources to be coordinated. This unified time-series monitoring baseline serves as the time reference framework for the entire resource pool coordination process. By writing synchronization timestamp records into the blockchain ledger, it achieves time-series alignment throughout the resource allocation process. Specifically, by calling the blockchain's timestamp function and an external trusted time source interface, the time offsets of different nodes are self-calibrated and corrected using linear regression, ensuring millisecond-level synchronization accuracy is maintained even in a multi-node asynchronous environment. Simultaneously, to improve the dynamic stability of the time-series baseline, an automatic time-series update process is triggered each time a resource request is submitted to the blockchain network, thereby enabling the unified time-series monitoring baseline to adaptively track the global time state.

[0022] After establishing a unified time-series monitoring baseline, the trigger links of smart contracts used for resource allocation in the blockchain are layered and labeled. Layered labeling refers to dividing the smart contracts into three levels—the basic trigger layer, the intermediate logic layer, and the strategy execution layer—based on their call hierarchy and dependencies in the resource scheduling logic. The basic trigger layer is responsible for on-chain registration and preliminary verification of resource request events; the intermediate logic layer handles resource competition, matching, and scheduling decisions; and the strategy execution layer completes the final resource locking and release operations. When labeling these levels, a combination of hash indexes and dependency matrices is used to generate a unique link identifier for each level. These identifiers are then associated with and stored using the timestamps of the unified time-series monitoring baseline, forming a dual-index system of hierarchy and time. This system allows for precise tracing and reproduction of call behavior at any time and level during subsequent analysis, providing structured data support for modeling recursive risks.

[0023] After completing the hierarchical labeling of the smart contract triggering chain, the call depth and latency parameters of each level are collected in real time based on the synchronized time data of the unified time series monitoring baseline. The call depth parameter describes the nesting level and recursion level of the smart contract chain, while the latency parameter reflects the response time changes of the call chain during actual execution. Specifically, by recording high-precision timestamps at the beginning and end of each call and calculating the difference to obtain the real-time latency, a two-dimensional dynamic matrix of call depth and latency is constructed by combining it with call depth statistics. To ensure the continuity and authenticity of data collection, all time and depth data are written to the blockchain ledger in the form of on-chain event logs, which are tamper-proof and traceable. Furthermore, by introducing a time series analysis algorithm based on big data artificial intelligence, dynamic sliding window calculations are performed on the real-time collected data to identify abnormal latency fluctuations or sudden increases in depth within specific time intervals, thereby capturing potential recursive stacking risk signals.

[0024] After obtaining complete call depth and latency parameter data, a recursive risk map containing dynamic constraints is generated based on the unified time-series monitoring baseline and hierarchical labeling structure. The generation process of the recursive risk map includes three key stages: multi-dimensional data fusion, temporal clustering, and constraint weight calculation. First, the call depth parameters, latency parameters, and corresponding timestamps at each level are mapped in a unified coordinate space to form a multi-dimensional time evolution curve. Then, using a temporal clustering algorithm based on dynamic adjustment of cluster centers, call links exhibiting synchronous fluctuations within adjacent time periods are aggregated into a risk cluster. Each risk cluster corresponds to a potential recursive association pattern, representing a combination of call links that may lead to deadlock or performance degradation. Further, based on the depth distribution characteristics and latency gradient of each link within a risk cluster, its dynamic constraint weights are calculated, forming a set of risk assessment parameters that can change over time. These parameters not only describe the intensity of the recursive risk but also serve as initial weight settings when introducing the force field calculation mechanism in subsequent steps. Finally, all risk clusters and their dynamic constraint weights are stored in a graph structure to form the recursive risk map. This recursive risk graph uses a unified time-series monitoring baseline as the time dimension, a hierarchically labeled smart contract triggering chain as the structural dimension, and dynamic constraint weights as the risk measurement dimension, thereby realizing the visual expression and quantitative evaluation of the recursive behavior of smart contracts.

[0025] Through the above steps, the entire process from time synchronization, hierarchical marking, data collection to risk modeling is completed before the blockchain resource pool coordination begins, providing a precise quantitative basis for subsequent anomaly detection, deadlock identification, and link optimization. This not only introduces a unified time-series monitoring baseline as a time reference mechanism in blockchain resource scheduling scenarios, but also transforms complex contract call relationships into a measurable risk graph through the fusion of hierarchical marking and dynamic constraints. This enables subsequent big data and artificial intelligence algorithms to perform self-learning and predictive optimization based on this graph, significantly improving the transparency, controllability, and intelligence level of blockchain resource coordination.

[0026] Step 2: Under the constraints of the recursive risk graph, a force field calculation mechanism is introduced to inject micro-scale perturbation signals into the smart contract triggering link to calculate the stability factor of the calling chain. Based on the calculation results, local unstable segments are extracted and solidified to form potential deadlock fingerprints, so as to achieve the initial identification of recursive risks. The specific implementation steps are as follows: This paper describes the initialization process of the force field calculation mechanism for smart contract trigger links used for resource allocation in a blockchain, under the constraints of a recursive risk graph. This initialization process is based on the dynamic constraint weights and hierarchical structure of the previously established recursive risk graph. Each smart contract trigger link is treated as a force-bearing unit, and a state evolution equation is established under a unified time-series monitoring baseline. Specifically, for each link node with a high-risk weight in the recursive risk graph, its force components are defined as three parts: time delay force, call depth force, and resource competitiveness. The time delay force characterizes the hindering effect of the link during execution, the call depth force reflects the complexity of the recursive nesting hierarchy, and the resource competitiveness reflects the degree of competition for on-chain resources by concurrent tasks. By introducing a force field function into the time coordinate of the unified time-series monitoring baseline, the above three force components are superimposed, and a dynamic force field distribution map is constructed in the multi-dimensional time domain. This allows the stability state of each contract trigger link at a given time to be quantitatively described by the balance of the force field distribution, laying the foundation for subsequent stability calculations involving injected perturbation signals.

[0027] After initializing the force field calculation mechanism, microscale perturbation signals are introduced to verify the dynamic response characteristics of high-risk link regions marked in the recursive risk graph. The design of this perturbation signal follows the principle of "non-destructive dynamic perturbation," meaning that it applies extremely small-amplitude periodic perturbations to the link's latency, resource contention, and state transition rate parameters without altering the contract logic execution results. By setting the perturbation period under a unified time-series monitoring baseline, the perturbation frequency is synchronized with the block generation time, thus avoiding analysis errors caused by external time-series deviations. After each round of perturbation signal injection, the system monitors the link's response changes in real time under the constraints of the recursive risk graph, collecting data such as time delay offset, call depth fluctuation, and resource contention intensity changes within the perturbation period. This data allows for further calculation of the link's local response sensitivity, serving as an important input for subsequent stability factor calculations. It is worth noting that the microscale perturbation signal is not injected all at once, but is continuously iterated according to the dynamic changes of the risk map, so that the calculation results of each round can reflect the latest link status, thus giving the force field calculation results temporal continuity and dynamic adaptability.

[0028] After completing the disturbance signal injection and response data acquisition, the stability factor of the call chain is calculated based on the obtained dynamic force field distribution data. The stability factor characterizes the dynamic equilibrium capability of the link under external disturbances; the closer its value is to the critical threshold, the closer the link is to instability or recursive deadlock. The specific calculation method is as follows: First, for each level of link node, the direction change rate of its resultant force vector is calculated based on its corresponding time delay force, call depth force, and resource competitiveness. Second, the dynamic response curve of the link is obtained by integrating the average amplitude of this change rate over a disturbance period. Then, a nonlinear fitting algorithm based on big data artificial intelligence is used to fit the amplitude and phase of the dynamic response curve, extracting its maximum response amplitude and stable convergence time as two core parameters. Finally, these two parameters are input into the stability calculation model, combined with the dynamic constraint weights of the corresponding nodes in the recursive risk map, to obtain the stability factor value of the link at the current moment. By calculating the changing trend of the stability factor within a continuous time window, link regions exhibiting periodic decay or sudden growth can be identified; the former indicates that the link is trending towards stability, while the latter foreshadows potential instability risks.

[0029] After obtaining the complete distribution of stability factors, local regions exhibiting significant instability are extracted and solidified to form potential deadlock fingerprints. This process includes three consecutive actions: identification of locally unstable segments, feature vector extraction, and fingerprint solidification and storage. First, based on the time-series distribution of stability factors, link segments whose stability factors exceed a set threshold within multiple consecutive time windows are selected as locally unstable segments. Second, the latency change rate, depth fluctuation amplitude, and resource contention gradient of these segments are normalized and recorded in the form of multi-dimensional feature vectors. Third, through the constraints of the recursive risk map, the spatial similarity of these feature vectors in the map is calculated. Link groups with similarity exceeding a set threshold are aggregated into a set of potential deadlock patterns, and these patterns are hashed to generate unique potential deadlock fingerprints. This potential deadlock fingerprint not only preserves the dynamic behavior characteristics of the links within a specific time period but also achieves cross-time-domain correlation through a unified time-series monitoring baseline. This allows for rapid location and reproduction of the formation trajectory of problematic links during subsequent anomaly identification and deadlock prevention. In this way, the entire recursive risk analysis process is transformed from passive detection to active identification, significantly improving the ability to perceive and respond to potential recursive risks in the coordination of blockchain resource pools.

[0030] By introducing a force field calculation mechanism under the constraint of a recursive risk graph, and combining it with microscale perturbation signals and stability factor calculations, accurate modeling and early risk identification of the dynamic characteristics of smart contract triggering chains are achieved. A dynamic stability assessment method based on force field analysis is introduced into the blockchain resource scheduling scenario, and unstable behavior characteristics are solidified in the form of potential deadlock fingerprints. This enables the quantification and tracking of complex contract call recursive relationships, providing a foundation for subsequent abnormal recursive chain modeling and adaptive decoupling.

[0031] Step 3: Based on the deadlock fingerprint, run the counterfactual replay chain to replay the historical contract triggering trajectory in the blockchain in a time sequence, strip away the normal call path and amplify the abnormal superposition link to generate an abnormal recursion chain model with causal traceability, so as to determine the formation mechanism of abnormal recursion and intervention strategy. The specific implementation steps are as follows: After obtaining the potential deadlock fingerprint, a counterfactual replay chain is initiated to replay the historical smart contract triggering trajectories in the blockchain in a temporal sequence. This counterfactual replay chain is built upon the local instability characteristics recorded by the potential deadlock fingerprint generated in the previous stage. By selecting historical block sequences related to the deadlock fingerprint within the time coordinates of a unified time-series monitoring baseline, the chain completely replays the input parameters, execution state, output results, and intermediate state transitions of each smart contract call. To ensure the accuracy of the replay, the counterfactual replay chain not only reproduces the original transaction execution logic but also introduces a virtual intervention parameter channel to simulate link behavior under different intervention conditions without affecting the actual state of the blockchain ledger. Specifically, by decoupling and rearranging the execution order of the original contract call trajectory in time, multiple possible historical execution paths can be replayed in parallel on the timeline, facilitating the identification of anomalous links that only lead to recursive superposition under specific triggering conditions. The core of this process lies in constructing an execution environment consistent with the actual historical state, ensuring that each counterfactual replay is synchronized under a unified time-series monitoring baseline, thereby ensuring the accuracy of subsequent anomalous path stripping and amplification analysis.

[0032] After the initial timing replay of the counterfactual replay chain, a process of stripping normal call paths begins for unstable link regions identified by potential deadlock fingerprints. This process analyzes the time evolution curves of call depth and latency parameters in each round of replay results, filtering out paths that meet the stability factor below a set threshold and whose execution results conform to the expected state transition pattern. These paths are defined as normal call paths. To prevent normal paths from being misjudged as abnormal paths, dynamic constraint weights from the recursive risk graph are introduced as correction factors during the stripping process. That is, each call path is weighted and scored according to its risk weight in the graph, with paths with lower weights being prioritized for identification as normal paths. After stripping, the system automatically marks the remaining paths as suspected abnormal links. These links typically exhibit abnormal growth in call depth, sudden increases in latency, or backtracking reentrancy behavior in multi-layered dependencies. Through this process, the counterfactual replay chain successfully decomposes the complex historical execution trajectory into two independent sets of paths, allowing subsequent analysis to focus on the truly abnormal superimposed links that may trigger implicit deadlocks.

[0033] After stripping away the normal path, anomaly superposition and amplification analysis is performed on the remaining suspected abnormal links to reveal the formation process of recursive anomalies. This analysis, based on multi-dimensional execution records in the counterfactual replay chain, utilizes time inversion and event dependency reconstruction methods to unfold and causally align the call sequence of the abnormal links layer by layer. Specifically, firstly, based on the time series of the unified time-series monitoring baseline, the trigger time, execution order, and result state of each abnormal link node are aligned to construct a time-causal matrix. Then, by introducing a stability factor defined in the force field calculation mechanism, dynamic weights are applied to each node in the matrix, making nodes with poor stability more significant in the causal chain. Next, using a time-series clustering algorithm based on big data artificial intelligence, the nodes in the causal matrix are grouped by similarity to identify node combinations with strong dependencies and repetitive recursive characteristics. Finally, these node combinations are reconstructed into a time evolution diagram of the abnormal superposition links according to their chronological order. This evolution diagram not only shows the propagation path of the abnormal recursion but also quantitatively describes the trend of recursion intensity over time, making the amplification process of the abnormal links visible and measurable. In this way, the root cause of the link structure that causes the resource pool coordination anomaly can be accurately revealed without interrupting the normal operation of the blockchain.

[0034] After completing the amplified analysis of anomalies, an anomaly recursive chain model with causal traceability is generated based on the amplified anomaly link data to determine the formation mechanism and intervention strategies for anomaly recursion. This model is constructed based on the time evolution graph of the anomaly superimposed links and formally models the causal relationships of all anomaly nodes under a unified time-series monitoring baseline. First, each anomaly node is considered a causal event unit, with its input being the pre-trigger state and its output being the post-trigger state, and a state transition function is established between nodes. Second, based on the dependencies in the time-causality matrix, the propagation probability and influence intensity between event units are calculated, thus forming a recursive propagation graph. Then, by introducing counterfactual simulation paths into this recursive propagation graph, the propagation results under different intervention variables (such as adjusting the execution order, changing resource priorities, or limiting the number of re-entries) are compared to extract the key nodes that can effectively block anomaly propagation. Finally, the control conditions and corresponding intervention strategies of these key nodes are solidified into the anomaly recursive chain model, enabling the model to not only describe the formation mechanism of anomaly recursion but also provide real-time intervention suggestions during subsequent resource coordination. Through this model, managers of blockchain resource pools can predictively prevent and control potential recursive anomalies during system operation, achieving closed-loop control from anomaly identification to intervention decisions.

[0035] Through the above steps, based on the potential deadlock fingerprint, a counterfactual replay chain was used to achieve the traceable reenactment of historical smart contract trigger trajectories. Furthermore, through normal path stripping and anomaly amplification analysis, an anomaly recursive chain model with causal reasoning capabilities was generated. By combining counterfactual deduction with blockchain execution timing and simulating state evolution under different historical conditions, the causal mechanism of hidden deadlocks was revealed. Based on this, a self-explanatory and self-learning anomaly recursive chain model was constructed, enabling the system to generate targeted intervention strategies for different anomaly types, thereby significantly improving the intelligence level of blockchain resource pool coordination.

[0036] Step 4: Based on the abnormal recursive chain model, construct an adaptive decoupling mapping, unlock the call chain that deviates from the normal state step by step and redirect it to a safe path. At the same time, write the correction results back to the unified time-series monitoring baseline, thereby realizing the dynamic repair and global balance recovery of the resource allocation channel. The specific implementation steps are as follows: After obtaining an anomaly recursive chain model with causal traceability, the adaptive decoupling mapping construction process is initiated based on the anomaly propagation path and key influencing nodes identified in the model. This process uses the anomaly recursive chain model as the input data source and combines time series information from a unified time series monitoring baseline to perform multi-dimensional feature analysis on call chains that deviate from the normal state. Specifically, an anomaly propagation matrix is ​​formed by extracting the state transition function, propagation probability, and influence intensity parameters of each node in the anomaly recursive chain model. Subsequently, based on the causal dependencies in the propagation matrix, a phase space mapping of the link states is established, projecting the dynamic behavior of the call chain from the time domain to the state space, thereby achieving an explicit distinction between anomaly states and steady states. In this phase space, the system calculates the deviation magnitude of each call chain according to the stability factor and dynamic constraint weights, and uses the deviation magnitude as the initial input of the adaptive decoupling mapping, enabling the subsequent unlocking and redirection processes to accurately identify the start point, end point, and evolution trend of the anomaly chain. The key to this stage is to transform the implicit causal structure in the anomaly recursive chain model into a computable mapping structure, thereby providing a quantitative basis for subsequent link unlocking and path redirection.

[0037] After constructing the initial structure of the adaptive decoupling mapping, the unlocking process of the deviated links is executed level by level according to the degree of deviation of the nodes in the anomaly propagation matrix. This process is guided by causal priority and unfolds layer by layer according to the time sequence from the source node to the terminal node in the anomaly recursive chain model. Specifically, under the time coordinate of the unified time-series monitoring baseline, the system first performs a state unlocking operation on the source node with the largest deviation, that is, temporarily freezes its triggering permission to downstream nodes to prevent the abnormal state from continuing to propagate; then, using the state similarity judgment rules in the decoupling mapping, the state transition paths of adjacent nodes are recalculated, and paths with similarity below a preset threshold are marked as anomaly link candidate set; next, the path with the lowest propagation probability is selected as the priority unlocking target from the candidate set, and its call depth parameter and resource competition priority are adjusted to remove it from the anomaly propagation chain, and a corresponding state change record is generated on the blockchain ledger. With the completion of each layer of unlocking operation, the system updates the parameters of the adaptive decoupling mapping in real time, so that the unlocking process of subsequent layers can automatically refer to the previous adjustment results, forming a progressive adaptive optimization mechanism. In this way, the system can gradually unlock the logic of the recursive exception chain without breaking the original contract logic, thus achieving a dynamic transition from local exceptions to global stability.

[0038] After unlocking the abnormal links step by step, a path redirection operation is performed on the identified and corrected link states. The goal of this operation is to redirect the call links that have escaped the abnormal state to a safe path, ensuring the continuity and consistency of the resource allocation process. Specifically, within the state space of the adaptive decoupling mapping, the time alignment deviation of each unlocked link is first calculated based on the time series information of the unified time-series monitoring baseline and the hierarchical markers in the recursive risk graph. Then, a dynamic path selection algorithm is used to find the steady-state path in the historical contract execution trajectory that is closest to the current link state, serving as the redirection target path. To prevent local optima in path selection, the system adopts a multi-objective optimization-based criterion, adjusting the weights between minimizing latency, maximizing stability, and balancing resource usage to ensure that the redirected path has optimal performance globally. After path redirection, the system records the path replacement event in the blockchain ledger and maps this event to the original abnormal node in the abnormal recursive chain model, achieving causal traceability from the abnormal state to the steady state. Through this mechanism, every anomaly repair in the resource pool is verifiable and traceable, thereby improving the transparency and credibility of the entire resource scheduling process.

[0039] After completing link unlocking and path redirection, the correction results are written back to the unified time-series monitoring baseline, thereby achieving dynamic repair and global equilibrium restoration of the resource allocation channel. The core of this process lies in ensuring the system's time-series consistency and data traceability. Specifically, the system first inserts a set of correction markers into the time coordinates of the unified time-series monitoring baseline to record the time point and adjustment magnitude of each link state adjustment. Then, the corrected call depth parameters, latency parameters, and resource contention priorities are synchronously updated to the recursive risk graph, ensuring that the dynamic constraints of the risk graph reflect the latest system state in real time. Next, using big data and artificial intelligence time-series prediction algorithms, trend analysis is performed on the updated baseline data to predict potential future recursive fluctuations and generate early risk warning signals. Finally, by recalculating resource allocation priority weights globally and adjusting the triggering order of smart contracts at each level, cross-level global equilibrium restoration is achieved. Throughout this process, all correction data and write-back records are written to the blockchain ledger in an immutable manner, forming a verifiable repair history, ensuring auditable traceability of every dynamic repair step coordinated by the resource pool. Through this mechanism, the system can not only quickly heal itself after an anomaly occurs, but also maintain dynamic balance during long-term operation, achieving continuous evolution from local link repair to overall resource pool balancing.

[0040] By constructing an adaptive decoupling mapping based on an abnormal recursive chain model, and combining continuous processes such as step-by-step unlocking, path redirection, and time-series write-back, self-repair and global equilibrium recovery of blockchain resource pool coordination are achieved. By combining a causal reasoning model with dynamic decoupling mapping and performing verifiable state write-back under a time-series monitoring baseline, the transformation of blockchain resource scheduling from "passive repair" to "active balancing" is realized. Simultaneously, through a path redirection and risk graph linkage mechanism, the system possesses self-organizing and self-evolving capabilities, thereby maintaining the continuity, stability, and optimality of resource allocation in complex, multi-source, heterogeneous environments.

[0041] This invention establishes a unified time-series monitoring baseline before resource pool coordination and combines a recursive risk graph and force field calculation mechanism to achieve real-time quantitative modeling and dynamic stability assessment of smart contract call chains during blockchain resource allocation. Through hierarchical labeling, time-series data acquisition, and micro-scale perturbation calculation, the latency fluctuations and recursion depth of each call chain can be continuously monitored, enabling early identification and location of potential deadlocks before anomalies spread. This mechanism not only effectively avoids resource locking and channel freezing caused by long-term call chain stacking but also significantly improves the scheduling transparency and response agility of the resource pool in multi-source heterogeneous environments, transforming the resource coordination process from static monitoring to an intelligent mode of proactive defense and adaptive prediction.

[0042] This invention achieves causal tracing, path repair, and global equilibrium restoration of abnormal recursive chains by introducing a counterfactual replay chain and adaptive decoupling mapping after anomaly identification. Through temporal replay of historical contract trajectories and amplified analysis of abnormal links, combined with adaptive unlocking and secure path redirection mechanisms, recursive deadlocks can be dynamically resolved and resource liquidity restored without interfering with the normal operation of the blockchain. The repair results are synchronously written back to a unified temporal monitoring baseline, enabling the system to have continuous self-correction capabilities, forming a closed-loop anomaly detection-intervention-repair mechanism. This technology not only improves the continuity of resource allocation and the robustness of system operation but also constructs a cross-level, traceable, and verifiable resource coordination system, effectively enhancing the global stability and long-term operational reliability of the blockchain resource pool.

[0043] This invention provides, for example Figure 2 The resource pool coordination system shown is based on blockchain and big data AI, including a time-series monitoring baseline construction module, a force field calculation and analysis module, a counterfactual playback chain modeling module, and an adaptive decoupling and repair module; The timing monitoring baseline construction module establishes a unified timing monitoring baseline before resource pool coordination begins. It hierarchically marks the smart contract triggering links used for resource allocation in the blockchain, collects the depth parameters and latency parameters of each level of call in real time, and generates a recursive risk map containing dynamic constraints based on the collection results. The force field calculation and analysis module, under the constraint of the recursive risk graph, introduces a force field calculation mechanism to inject micro-scale perturbation signals into the smart contract triggering link to calculate the stability factor of the calling chain, and extracts local unstable sections based on the calculation results, and solidifies the sections to form potential deadlock fingerprints. The counterfactual replay chain modeling module runs a counterfactual replay chain based on deadlock fingerprints, replays the historical contract triggering trajectory in the blockchain in a time sequence, strips out the normal call path and amplifies the abnormal superimposed link, and generates an abnormal recursive chain model with causal traceability. The adaptive decoupling repair module constructs an adaptive decoupling mapping based on the abnormal recursive chain model, unlocks and redirects call chains that deviate from the normal state to safe paths step by step, and writes the correction results back to the unified timing monitoring baseline.

[0044] The resource pool coordination method based on blockchain and big data AI provided in this embodiment of the invention is implemented through the above-mentioned resource pool coordination system based on blockchain and big data AI. For details of the specific methods and processes of the resource pool coordination system based on blockchain and big data AI, please refer to the embodiment of the resource pool coordination method based on blockchain and big data AI, which will not be repeated here.

[0045] The foregoing has only described certain exemplary embodiments of the present invention by way of illustration. Undoubtedly, those skilled in the art can modify the described embodiments in various ways without departing from the spirit and scope of the present invention. Therefore, the foregoing drawings and descriptions are illustrative in nature and should not be construed as limiting the scope of protection of the claims of the present invention.

Claims

1. A resource pool coordination method based on blockchain and big data AI, characterized in that, Includes the following steps: Step 1: Before resource pool coordination begins, establish a unified time-series monitoring baseline, hierarchically mark the smart contract triggering links used for resource allocation in the blockchain, collect the depth parameters and latency parameters of each level of call in real time, and generate a recursive risk map containing dynamic constraints based on the collection results. Step 2: Under the constraints of the recursive risk graph, a force field calculation mechanism is introduced to inject microscale perturbation signals into the smart contract triggering link to calculate the stability factor of the calling chain, and extract local unstable sections based on the calculation results, and solidify the sections to form potential deadlock fingerprints. Step 3: Based on the deadlock fingerprint, run the counterfactual replay chain to replay the historical contract triggering trajectory in the blockchain in a time sequence, strip away the normal call path and amplify the abnormal superimposed link to generate an abnormal recursive chain model with causal traceability. Step 4: Construct an adaptive decoupling mapping based on the abnormal recursion chain model, unlock the call chain that deviates from the normal state step by step and redirect it to the safe path, and write the correction results back to the unified timing monitoring baseline.

2. The resource pool coordination method based on blockchain and big data AI according to claim 1, characterized in that, The steps to establish a unified time-series monitoring baseline include: Before resource pool coordination begins, the timestamp function of the blockchain ledger is called and combined with the external trusted time source interface to perform self-calibration and linear regression correction on the time offset of multiple nodes, and establish a unified time series monitoring baseline for global time alignment. After establishing a unified time-series monitoring baseline, the triggering links of smart contracts used for resource allocation in the blockchain are hierarchically marked, and hierarchical identifiers are generated by combining hash index and dependency matrix, so that they are associated with and stored with the timestamp of the unified time-series monitoring baseline. Based on the synchronization time data of the unified time series monitoring baseline, the call depth parameters and latency parameters of the smart contract triggering chain at each level are collected in real time, and a two-dimensional dynamic matrix of call depth-latency is constructed. Based on the dynamic changes of the matrix, a time-series clustering algorithm is used to calculate the dynamic constraint weights of risk clusters, generating a recursive risk map with a unified time-series monitoring baseline as the time dimension.

3. The resource pool coordination method based on blockchain and big data AI according to claim 2, characterized in that, The steps for generating a recursive risk map include: when calculating the dynamic constraint weights of risk clusters, based on the call depth parameters and latency gradients of the trigger links of smart contracts at each level, using a time-series clustering algorithm with dynamically adjusted cluster centers to aggregate synchronously fluctuating links, obtaining time-series weight parameters that can reflect changes in the intensity of recursive risk, and using these weight parameters as the initial input values ​​for the subsequent force field calculation mechanism.

4. The resource pool coordination method based on blockchain and big data AI according to claim 1, characterized in that, The steps for introducing a force field calculation mechanism under the constraint of a recursive risk graph include: Based on the dynamic constraint weights and hierarchical structure in the recursive risk graph, each smart contract triggering link is regarded as a force-bearing unit, and a state evolution equation is established under a unified time-series monitoring baseline. After the force field calculation mechanism is initialized, microscale disturbance signals are injected into high-risk link areas and data on delay offset, call depth fluctuation and resource contention intensity changes are collected in real time. The stability factor of the link is calculated based on the collected data, and a dynamic response curve is generated. Locally unstable regions where the stability factor exceeds a threshold are extracted, and potential deadlock fingerprints are generated based on feature vector similarity to achieve initial identification of recursion risks.

5. The resource pool coordination method based on blockchain and big data AI according to claim 4, characterized in that, In the step of injecting microscale perturbation signals, the perturbation signals are non-destructive dynamic perturbation signals, and their injection frequency is synchronized with the block generation time to ensure that the latency parameters, resource contention parameters and state transition rate parameters of the link are periodically perturbed without affecting the execution result of the smart contract logic.

6. The resource pool coordination method based on blockchain and big data AI according to claim 1, characterized in that, The steps involved in running a counterfactual replay chain based on a deadlock fingerprint include: Based on the potential deadlock fingerprint, relevant historical block sequences are selected under a unified time-series monitoring baseline, and the input parameters, execution status, and intermediate state transition process of each smart contract call are replayed in time sequence. The replay results are analyzed using call depth and latency parameters. The dynamic constraint weights of the recursive risk graph are used to separate normal call paths and mark suspected abnormal links. Based on the execution time inversion and event dependency reconstruction of suspected abnormal links, a time-causality matrix is ​​constructed and a time evolution diagram of abnormal superimposed links is generated; Based on the time evolution diagram, an abnormal recursive chain model is established. Key nodes are extracted by calculating the transmission probability and influence intensity of event nodes, and corresponding intervention strategies are solidified.

7. The resource pool coordination method based on blockchain and big data AI according to claim 6, characterized in that, The steps for establishing an anomaly recursive chain model include: after generating the time evolution diagram of the anomaly superposition link, calculating the transmission probability and influence intensity of each anomaly node based on the time-causality matrix, comparing the propagation results under different intervention variables by introducing counterfactual simulation paths, identifying the key node that can effectively block the anomaly recursive propagation, and solidifying the control conditions and intervention strategies of the key node into the anomaly recursive chain model.

8. The resource pool coordination method based on blockchain and big data AI according to claim 1, characterized in that, The steps for constructing an adaptive decoupling map based on the abnormal recursion chain model include: Based on the anomaly propagation path and key impact nodes in the anomaly recursive chain model, an anomaly propagation matrix is ​​established by combining the time series information of the unified time series monitoring baseline, and a link state phase space mapping is constructed to distinguish between anomaly states and steady states. The deviated link is unlocked step by step according to the degree of node deviation, and the priority unlocking target is determined by state similarity judgment and propagation probability calculation; For unlocked links, perform path redirection, guide them to a safe path using a dynamic path selection algorithm, and record path replacement events; The correction results are written back to the unified time-series monitoring baseline and the recursive risk map is updated synchronously to achieve dynamic repair and global balance restoration of resource allocation channels.

9. A resource pool coordination system based on blockchain and big data AI, used to implement the resource pool coordination method based on blockchain and big data AI as described in any one of claims 1-8, characterized in that, It includes a time-series monitoring baseline construction module, a force field calculation and analysis module, a counterfactual playback chain modeling module, and an adaptive decoupling and repair module; The timing monitoring baseline construction module establishes a unified timing monitoring baseline before resource pool coordination begins. It hierarchically marks the smart contract triggering links used for resource allocation in the blockchain, collects the depth parameters and latency parameters of each level of call in real time, and generates a recursive risk map containing dynamic constraints based on the collection results. The force field calculation and analysis module, under the constraint of the recursive risk graph, introduces a force field calculation mechanism to inject micro-scale perturbation signals into the smart contract triggering link to calculate the stability factor of the calling chain, and extracts local unstable sections based on the calculation results, and solidifies the sections to form potential deadlock fingerprints. The counterfactual replay chain modeling module runs a counterfactual replay chain based on deadlock fingerprints, replays the historical contract triggering trajectory in the blockchain in a time sequence, strips out the normal call path and amplifies the abnormal superimposed link, and generates an abnormal recursive chain model with causal traceability. The adaptive decoupling repair module constructs an adaptive decoupling mapping based on the abnormal recursive chain model, unlocks and redirects call chains that deviate from the normal state to safe paths step by step, and writes the correction results back to the unified timing monitoring baseline.

10. A resource pool coordination device based on blockchain and big data AI, comprising the resource pool coordination system based on blockchain and big data AI as described in claim 9.

Images