Station early warning storage self-healing method and system based on peer edge mutual backup

Abstract

The application discloses a station early warning storage evidence self-healing method and system based on peer edge mutual backup, and relates to the technical field of data communication. The method comprises the following steps: S1, collecting multi-source operation data of edge nodes in real time, preprocessing, forming standardized early warning records, and constructing chain storage evidence abstract; S2, evaluating communication quality, determining the main mutual backup node and the candidate node list, and executing candidate takeover when the main node is unreachable; S3, performing alignment matching according to the chain storage evidence abstract, judging the consistency degree of the chain tail and the boundary offset degree, evaluating the boundary reliability, and realizing data completion and incremental recovery; S4, judging the connected state, executing data mutual backup, degradation adjustment and network interruption storage control, and triggering the self-healing process when the network is recovered. The problems that the confirmed data boundaries between nodes are inconsistent due to data confirmation delay during the main mutual backup node failure or switching process, and that the chain storage evidence structure appears repeated chain segments or implicit breakpoints are solved.

Description

Technical Field

[0001] This invention relates to the field of data communication technology, and in particular to a site early warning, evidence storage and self-healing method and system based on peer-to-peer edge redundancy. Background Technology

[0002] With the rapid deployment of new energy vehicle charging infrastructure and the continuous expansion of charging stations, along with the dense integration of charging equipment, battery operation status monitoring and safety early warning have become crucial foundations for ensuring operational safety and service quality. Charging stations typically use multi-source sensors to collect key parameters such as voltage, current, temperature, and insulation status in real time, and rely on edge computing and cloud platforms for data processing, early warning analysis, and operational management. In this process, early warning data not only serves as real-time safety monitoring but is also gradually becoming an important data basis for battery lifecycle management, fault tracing, liability determination, and compliance supervision.

[0003] For example, Chinese patent CN119945766A discloses a network security protection system and method for new energy power stations based on multimodal intelligent perception, which relates to the field of network security protection technology for new energy power stations. The system includes the following components: a multimodal data acquisition module, a data storage and management unit, a dynamic causal model construction and analysis module, a source tracing analysis and responsibility identification module, and a protection decision and execution module. Through dynamic causal model construction and analysis, this invention can update and reshape the causal network in real time to adapt to changes in the operating conditions of new energy power stations, enabling the invention to continuously learn and adapt to new network environments, thereby improving the intelligence level of the invention. By mining the potential causal relationships between multimodal data, it can promptly discover and warn of potential network security risks. At the same time, combined with source tracing analysis and responsibility identification, it can quickly locate the source of the event and assign responsibility to the responsible parties.

[0004] For example, Chinese patent CN119945772A discloses a network security monitoring and early warning method and system for new energy power plants, which relates to the field of network security technology. The method includes the following components: energy flow and information flow acquisition: In various equipment of new energy power plants, including but not limited to key parts of power generation equipment, power transformation equipment, energy storage equipment, and main energy transmission paths, this invention organically combines the collaborative analysis of energy flow and information flow with the establishment of a correlation model. It collects and analyzes energy flow and information flow data in real time, establishes the correlation between the two, and makes a comprehensive judgment based on the real-time status of energy flow when information flow is abnormal. This effectively identifies network attacks aimed at interfering with energy flow output, thereby comprehensively and accurately perceiving potential network security threats. In addition, by monitoring the energy conversion efficiency of the equipment in real time and combining historical data and equipment operating conditions for analysis, the impact of network attacks on the equipment operating status can be discovered.

[0005] However, in a field environment where distributed edge nodes operate collaboratively, each edge node synchronizes and backs up early warning data through a primary-backup mechanism and dynamically selects a backup node to take over based on communication status. In this scenario, when the primary-backup node fails or becomes unreachable, the system switches to the backup node and performs incremental data synchronization. However, due to the uncertainty of the primary node's failure time and the delay in data confirmation, the "confirmed data boundaries" between the local node, the original primary node, and the backup node are inconsistent. This results in some early warning records potentially being synchronized repeatedly or not being fully received by any node, leading to chain segment duplication or hidden breakpoints. This problem is particularly prominent in high-frequency early warning generation scenarios, affecting the consistency of subsequent evidence storage chains.

[0006] Therefore, in order to address the above issues, there is an urgent need for a site early warning, evidence storage, and self-healing method and system based on peer-to-peer edge redundancy. Summary of the Invention

[0007] To address the technical problem in existing technologies where data confirmation delays during primary / backup node failures or switchovers lead to inconsistent boundaries of confirmed data between nodes, resulting in duplicate chain segments or hidden breakpoints in the chain-based evidence storage structure, this invention provides a site early warning evidence storage self-healing method and system based on peer-to-peer edge backup. The technical solution is as follows:

[0008] A self-healing method for site early warning and evidence storage based on peer-to-peer edge mutual backup is provided. This method includes: S1, real-time collection of multi-source operational data from edge nodes, time-series unification and validity screening of the multi-source operational data, formation of standardized early warning records based on early warning judgment, construction of a continuous chain-like evidence storage sequence, and extraction of chain-like evidence storage digests; S2, communication quality assessment based on inter-node connectivity, determination of primary and backup node lists and candidate node lists based on the communication quality assessment, establishment of data mutual backup relationships, and execution of candidate takeover and takeover request transmission when the primary node is unreachable; S3, for the takenover node, alignment matching is performed based on the chain-like evidence storage digest, determination of chain tail consistency and boundary offset, formation of boundary reliability assessment results, and identification of the data recovery start position and recovery interval based on the boundary reliability assessment results, achieving data completion and incremental recovery; S4, connectivity status determination based on heartbeat interaction results, execution of data mutual backup, degradation adjustment, and network outage evidence storage control under different connectivity states, and triggering a self-healing process when the network recovers, completing chain-like data consistency recovery and continuous evidence storage reconstruction.

[0009] Furthermore, real-time multi-source operational data from edge nodes is collected, and the multi-source operational data undergoes time-series unification and validity screening. Based on early warning judgments, standardized early warning records are formed, and a continuous chain-like evidence storage sequence is constructed. The specific process for extracting the chain-like evidence storage summary is as follows: Edge nodes collect raw operational data in real-time based on the sensor modules of the corresponding charging pile locations. The raw operational data includes: voltage, current, temperature, insulation resistance, node reception time, and the node's unique identifier. Using the node reception time as a unified time reference, time-unified processing is performed on each piece of raw operational data collected, and records are made according to a fixed sampling period. For any edge node, if there are two or more raw operational data records within the fixed sampling period, only the record with the earliest node reception time is retained, and the remaining raw data records are not recorded. The system executes data and determines whether the original operational data is within a physically valid range. If not, the corresponding original operational data record is deleted. For each original operational data record, a warning record is defined when the temperature exceeds the temperature warning threshold or the current exceeds the current warning threshold. Warning records are sorted in ascending order according to the node reception time within the same edge node, and a continuously increasing warning record number is generated. The warning record number and the corresponding original operational data are combined to form a standardized warning record. The current standardized warning record is concatenated with the previous chained evidence value according to the order of the warning record numbers within the same edge node, and then input into a hash function to calculate the current chained evidence value. A chained evidence value sequence of a fixed length is traced back from the current maximum warning record number to serve as the chained evidence digest.

[0010] Furthermore, based on the connectivity between nodes, communication quality is assessed. The specific process for determining the primary backup nodes and candidate nodes and establishing data backup relationships is as follows: For any two edge nodes, at least N connectivity probe messages are sent within a fixed measurement period, and the round-trip delay for each probe is recorded. The minimum round-trip delay is taken as the actual communication delay between the node pairs. The minimum actual communication delay among all node pairs is taken as the globally optimal delay. The globally optimal delay is divided by the actual communication delay of each node pair to obtain the communication quality of the node pairs. For each edge node, the communication quality with all other edge nodes is calculated, and communication is selected... The edge node with the highest quality is designated as the primary backup node for the current edge node. Edge nodes are then sorted by communication quality from highest to lowest, and the top M edge nodes ranked after the primary backup node are selected as candidate nodes. A data backup relationship is established between the current edge node and its corresponding primary backup node. Under this relationship, after generating each standardized warning record, the current edge node sends the standardized warning record, its corresponding warning record number, and the chained evidence value to the primary backup node. Upon receiving the standardized warning record, the primary backup node stores it in sequence according to the warning record number and calculates the corresponding chained evidence value for each received standardized warning record, performing consistency verification.

[0011] Furthermore, the specific process of executing the backup takeover and sending the takeover request when the primary node is unreachable is as follows: For each edge node, a heartbeat message is sent to the primary backup node according to a fixed heartbeat detection period, and the heartbeat sending time and response receiving time are recorded. When no heartbeat response is received from the corresponding primary backup node within K consecutive heartbeat detection periods, the primary backup node is determined to be in an unreachable state. When the primary backup node is in an unreachable state, the edge node with the highest communication quality is selected from the candidate node list as the new primary backup node, and a takeover request message is sent to the new primary backup node. The takeover request message contains the maximum warning record sequence number of the current edge node, the chained evidence digest, and the node's unique identifier.

[0012] Furthermore, for the node under takeover, the specific process of alignment and matching based on the chain-based evidence digest is as follows: The new primary backup node receives and parses the takeover request message. Simultaneously, the new primary backup node reads the standardized early warning record sequence of the corresponding edge node that it has already stored, determines the local maximum early warning record number, and calculates the chain-based evidence value one by one based on the stored standardized early warning records to form a local chain-based evidence value sequence; it backtracks from the local maximum early warning record number to records of the same length as the chain-based evidence digest, extracts the corresponding chain-based evidence value sequence, and forms a local chain-based evidence digest; and then sets the current edge node... The chained evidence digest of the edge node is aligned and matched with the local chained evidence digest. Starting from the end value of the local chained evidence digest, the matching position with the same chained evidence value and closest to the end of the chained evidence digest in the current edge node's chained evidence digest is used as the alignment position. Starting from the alignment position, the comparison is performed item by item backward. When the corresponding chained evidence values ​​are equal, the count is incremented by one. When they are not equal, the comparison is stopped, and the length of the longest consecutive identical suffix is ​​obtained. The round-trip time of the historical connectivity probe messages between the current edge node and the new primary backup node is calculated, and the mean and standard deviation of the round-trip time are calculated.

[0013] Further, the specific process for determining the chain tail consistency degree and boundary offset degree to form the boundary reliability assessment result is as follows: Divide twice the length of the longest consecutive identical suffix by the sum of the chain-based evidence digest length sent by the current edge node and the local chain-based evidence digest length to obtain the chain tail overlap ratio; calculate the absolute difference between the maximum warning record number of the current edge node and the maximum warning record number locally, and take the minimum value between the maximum warning record number of the current edge node and the maximum warning record number locally. Divide the absolute difference by the sum of the minimum value and the smallest positive number to obtain the relative boundary offset. Take the negative value of the relative boundary offset and perform natural exponentiation to obtain the boundary offset suppression term; divide the round-trip delay standard deviation by the sum of the round-trip delay mean and the smallest positive number to obtain the delay dispersion ratio. Add one to the obtained delay dispersion ratio and take the reciprocal to obtain the delay stability term; multiply the chain tail overlap ratio term, the boundary offset suppression term, and the delay stability term together to obtain the boundary alignment value.

[0014] Furthermore, based on the boundary reliability assessment results, the specific process of identifying the data recovery start position and recovery interval, and realizing data completion and incremental recovery is as follows: Multiply the difference between the boundary alignment value and the longest consecutive identical suffix length by the length of the longest consecutive identical suffix, and round up to obtain the boundary rollback number. Subtract the longest consecutive identical suffix length from the maximum warning record number of the current edge node, add one, and then subtract the boundary rollback number to obtain the common data boundary number. When the local maximum warning record number is less than the common data boundary number, it is determined that there is a data gap before the common data boundary in the new primary backup node. Read the interval between the local maximum warning record number plus one and the common data boundary number as the missing data completion interval. The current edge node sends the standardized warning records within the missing data completion interval, the corresponding warning record numbers, and the chained evidence values ​​to the new primary backup node. After completing the missing data completion, read the common data boundary number... The interval between the local maximum warning record number plus one and the current edge node's maximum warning record number is used as the incremental synchronization interval. Standardized warning records, their corresponding warning record numbers, and chained evidence values ​​within this interval are sent. When the local maximum warning record number is greater than or equal to the common data boundary number, it is determined that the new primary backup node has covered the data before the common data boundary. The interval between the local maximum warning record number plus one and the current edge node's maximum warning record number is read and used as the data synchronization interval. The current edge node sends the standardized warning records, their corresponding warning record numbers, and chained evidence values ​​within this data synchronization interval to the new primary backup node. After receiving the data, the new primary backup node performs storage according to the warning record number order, calculates the chained evidence value for each received standardized warning record, updates the local chained evidence value sequence, and returns a data synchronization completion confirmation message to the current edge node.

[0015] Further, the connectivity state is determined based on the heartbeat interaction result, and the specific processes of performing data mutual backup, degradation adjustment and offline evidence storage control under different connectivity states are as follows: within a fixed statistical time window, count the number of heartbeat packets sent by the current edge node to the current primary mutual backup node and the number of successfully received responses, divide the number of successfully received responses by the number of sent heartbeat packets to obtain the connectivity C between the current edge node and the primary mutual backup node, and compare the connectivity C with multi-level connectivity thresholds C1 and C2; when C≥C2, determine that the current edge node is in a normal connected state; when C1≤C<C2, determine that the current edge node is in a connected degradation state; when C<C1, determine that the current edge node is in an offline state; when the edge node is in a normal connected state, maintain the generation of standardized early warning records, the data mutual backup relationship between primary mutual backup nodes, and the calculation of chained evidence storage values; when the edge node is in a connected degradation state, maintain the generation of standardized early warning records, the data mutual backup relationship between primary mutual backup nodes, and the calculation of chained evidence storage values on the premise of shortening the heartbeat detection period; when the edge node is in an offline state, suspend sending standardized early warning records, early warning record sequence numbers and chained evidence storage values to the primary mutual backup node, and only continue to perform generation of standardized early warning records, update of chained evidence storage digest and calculation of chained evidence storage values locally.

[0016] Further, when the network recovers, a self-healing process is triggered, and the specific process of completing chained data consistency recovery and continuous evidence storage reconstruction is as follows: when the running state of the node recovers from the offline state or the connected degradation state to the normal connected state, trigger the offline self-healing recovery process: determine the current primary mutual backup node of the edge node, send the maximum early warning record sequence number, the corresponding chained evidence storage digest and the unique node identifier to the primary mutual backup node, calculate the boundary alignment value, complete the determination of the common data boundary sequence number, the judgment of the data missing interval and the data synchronization interval, and the supplementary transmission of the corresponding standardized early warning records, and update the standardized early warning record sequence and the chained evidence storage value sequence in the primary mutual backup node.

[0017] The second aspect of this invention provides a self-healing system for site early warning and evidence storage based on peer-to-peer edge redundancy, comprising: a site sensing data processing module, used to collect multi-source operational data from edge nodes in real time, perform time-series unification and validity screening on the multi-source operational data, form standardized early warning records based on early warning judgment, construct a continuous chain-like evidence storage sequence, and extract chain-like evidence storage summaries; and a peer-to-peer edge redundancy takeover module, used to perform communication quality assessment based on inter-node connectivity, determine a list of primary redundancy nodes and candidate nodes based on the communication quality assessment, establish data redundancy relationships, and execute candidate takeover and takeover requests when the primary node is unreachable. Request to send; Continuous evidence storage boundary recovery module, used for nodes under takeover, to perform alignment matching based on chain evidence storage digests, determine the consistency degree of the chain tail and the degree of boundary offset, form a boundary reliability assessment result, and identify the data recovery start position and recovery interval based on the boundary reliability assessment result, to realize data completion and incremental recovery; Evidence storage state switching self-healing module, used to determine the connectivity state based on heartbeat interaction results, perform data backup, degradation adjustment and network disconnection evidence storage control under different connectivity states, and trigger the self-healing process when the network is restored, to complete chain data consistency recovery and continuous evidence storage reconstruction.

[0018] The beneficial effects of the technical solutions provided in the embodiments of the present invention include at least the following:

[0019] (1) By performing time-series unification and validity screening on multi-source operational data and constructing a continuous chain of evidence storage sequence, this invention can ensure that the early warning data has structural consistency and traceability in the generation stage, effectively avoiding the problem of discontinuous evidence storage chain caused by data time sequence disorder or the introduction of abnormal data.

[0020] (2) This invention uses a dynamic selection and backup node selection mechanism based on communication quality to quickly switch nodes when the master node is unreachable, thereby avoiding the impact of single point failure on system operation and improving the stability and overall reliability of data backup between distributed nodes.

[0021] (3) Through the chain-based evidence storage digest alignment and boundary reliability assessment mechanism, the present invention can accurately identify the confirmed data boundaries between nodes, avoid repeated synchronization or missed synchronization problems during the data recovery process, thereby effectively eliminating chain segment duplication and hidden breakpoints, and ensuring the integrity and consistency of the chain-based evidence storage structure.

[0022] (4) This invention achieves continuous operation of local evidence storage in the event of network outage or connectivity degradation by using a heartbeat-based connectivity status determination and multi-state control strategy. After the network is restored, it automatically triggers a self-healing recovery process to achieve automatic data completion and chain reconstruction, which significantly enhances the robustness and self-recovery capability of the system in complex network environments. Attached Figure Description

[0023] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0024] Figure 1 This is a flowchart of the self-healing method for early warning and evidence storage at the site based on peer-to-peer edge mutual backup provided in this embodiment of the invention;

[0025] Figure 2 This is a structural diagram of the self-healing system for early warning and evidence storage at the site based on peer-to-peer edge redundancy provided in this embodiment of the invention;

[0026] Figure 3 This is a three-dimensional evaluation diagram of boundary alignment reliability based on multi-factor fusion provided in an embodiment of the present invention;

[0027] Figure 4 This is a flowchart of collaborative recovery process for network outage self-healing based on chain-based evidence storage and boundary alignment provided in an embodiment of the present invention. Detailed Implementation

[0028] The technical solution of the present invention will now be described with reference to the accompanying drawings.

[0029] In embodiments of the present invention, words such as "exemplarily," "for example," etc., are used to indicate that something is an example, illustration, or description. Any embodiment or design described as "exemplary" in the present invention should not be construed as being more preferred or advantageous than other embodiments or designs. Specifically, the use of the word "exemplary" is intended to present the concept in a concrete manner. Furthermore, in embodiments of the present invention, the meaning expressed by "and / or" can be both, or either one.

[0030] In the embodiments of this invention, the terms "image" and "picture" may sometimes be used interchangeably. It should be noted that, without emphasizing the distinction between them, they convey the same meaning. Similarly, the terms "of," "corresponding (relevant)," and "corresponding" may sometimes be used interchangeably. It should be noted that, without emphasizing the distinction between them, they convey the same meaning.

[0031] In this embodiment of the invention, sometimes a subscript such as W1 may be written in a non-subscript form such as W1. When the difference is not emphasized, the meaning they express is the same.

[0032] To make the technical problems, technical solutions and advantages of the present invention clearer, a detailed description will be given below in conjunction with the accompanying drawings and specific embodiments.

[0033] This invention provides a self-healing method for early warning and evidence storage at power stations based on peer-to-peer edge redundancy, such as... Figure 1 As shown, the processing flow of this method may include the following steps: S1, real-time collection of multi-source operation data of edge nodes, time-series unification and validity screening of multi-source operation data, formation of standardized early warning records based on early warning judgment, construction of continuous chain-like evidence storage sequence, and extraction of chain-like evidence storage digest; S2, communication quality assessment based on inter-node connectivity, determination of primary backup nodes and candidate node lists based on communication quality assessment, establishment of data backup relationship, and execution of candidate takeover and takeover request sending when the primary node is unreachable; S3, for the takenover node, alignment matching is performed based on the chain-like evidence storage digest, determination of chain tail consistency degree and boundary offset degree, formation of boundary reliability assessment result, and identification of data recovery start position and recovery interval based on boundary reliability assessment result, realizing data completion and incremental recovery; S4, connectivity status determination based on heartbeat interaction result, execution of data backup, degradation adjustment and network outage evidence storage control under different connectivity states, and triggering self-healing process when the network recovers, completing chain-like data consistency recovery and continuous evidence storage reconstruction.

[0034] Optionally, multi-source operational data from edge nodes is collected in real time. The multi-source operational data undergoes time-series unification and validity screening. Based on early warning judgments, standardized early warning records are formed, and a continuous chain-like evidence storage sequence is constructed. The specific process for extracting the chain-like evidence storage summary is as follows: Edge nodes collect raw operational data in real time based on the sensor modules of the corresponding charging pile locations. The raw operational data includes: voltage, current, temperature, insulation resistance, node reception time, and a unique node identifier. The sensor module includes a voltage acquisition unit, a current acquisition unit, a temperature acquisition unit, and an insulation detection unit. Each acquisition unit uploads data through a unified data acquisition interface and encapsulates it in a unified data format to ensure the structural consistency and resolvability of the multi-source data. The node reception time is the system timestamp of the edge node receiving the corresponding raw operational data, generated using the standard clock of the edge node's local clock. Using the node reception time as a unified time reference, time-uniform processing is performed on each piece of raw operational data collected. The node reception time is generated based on the local clock of the edge node, and the clocks of each edge node are aligned through a time synchronization mechanism. Network Time Protocol (NTP) is preferably used to achieve periodic time synchronization to reduce clock deviations between different nodes. Records are recorded according to a fixed sampling period. For any edge node, if there are two or more raw operational data records within the fixed sampling period, only the record with the earliest reception time is retained. For the remaining raw operational data, it is determined whether the raw operational data is within the physical valid range. If it is not, the corresponding raw operational data record is deleted. The fixed sampling period is a pre-set time interval, preferably 1 to 10 seconds, to ensure the consistency of data acquisition frequency. The physical valid range is determined based on the equipment operating specifications, specifically obtained through the equipment model parameter table written to the database during initialization. For example, voltage, current, and temperature correspond to preset upper and lower limit ranges, respectively. Data validity is filtered by judging whether the collected value falls within the corresponding range. The occurrence of two or more original operational data records within the same fixed sampling period is typically caused by sensor sampling jitter, communication delays leading to retransmissions, or inconsistent reporting times from multiple data sources. By retaining only the record with the earliest received time, a unified and stable time benchmark can be established, avoiding time ambiguity and duplicate calculations caused by multiple data entries within the same time window participating in subsequent processing. Furthermore, since the warning determination is based on threshold comparison and has continuous time window characteristics, the removal of redundant data within a single period will not substantially affect the overall warning trend identification, thus ensuring both time consistency and the stability and accuracy of the warning determination. For each original operational data record, a warning record is determined when the temperature exceeds the temperature warning threshold or the current exceeds the current warning threshold. The temperature and current warning thresholds are determined based on the equipment's rated parameters and can be dynamically adjusted according to the operating environment to improve the accuracy of the warning determination.For early warning records, they are sorted in ascending order according to the node reception time within the same edge node, and a continuously increasing early warning record number is generated. The early warning record number starts from an initial value of 1 and is unique within the same edge node, used to identify the generation order of the early warning records. The early warning record number and the corresponding original operational data are combined to form a standardized early warning record. The standardized early warning record is encapsulated using a unified data structure, including at least the early warning record number, the original operational data field, the node reception time, and the node's unique identifier, to facilitate subsequent storage and transmission. Following the sequence number of warning records within the same edge node, the current standardized warning record is concatenated with the previous chained evidence value and then input into a hash function to calculate the current chained evidence value. The concatenation order is "standardized warning record data plus the previous chained evidence value." The hash function preferably uses the SHA-256 one-way hash algorithm to ensure the immutability and security of the chained evidence structure. When the current warning record number is the initial value, the previous chained evidence value is a preset initial value, preferably a fixed-length string of zeros, and further preferably a string of all zero bits matching the output of the hash function. For example, when using SHA-256, the preset initial value is preferably a 256-bit all-zero value. A fixed-length chained evidence value sequence is traced back from the current maximum warning record number to serve as the chained evidence digest. The fixed length is the preset digest length, preferably 10 to 30 entries, used to reduce transmission and comparison overhead while ensuring the integrity of the digest information. The chained evidence digest is used for subsequent inter-node consistency verification and boundary alignment. Furthermore, after generating the chained evidence digest, each edge node continuously maintains a local continuous hash chain based on the current chained evidence value sequence. Specifically, for each new standardized early warning record generated, a corresponding chained evidence value is generated according to the early warning record sequence number, and the current chained evidence value is appended to the end of the local chained evidence value sequence to form a continuously increasing local hash chain structure, thereby achieving local continuous evidence storage. Each edge node can also send cloud anchor points to the cloud platform according to a preset anchoring period. The cloud anchor point includes the current maximum early warning record sequence number, the corresponding chained evidence value, the node's unique identifier, and the anchoring generation time. After receiving the cloud anchor point, the cloud platform establishes a cloud anchoring record for the corresponding edge node according to the node's unique identifier and stores it according to the anchoring generation time sequence to maintain the global continuous evidence storage time benchmark for the site's early warning data.

[0035] In this implementation plan, by performing unified time-series processing and validity screening on multi-source operational data, combined with early warning judgment and standardized encapsulation, the generation of early warning records with reliable data sources and unified structure is achieved. On this basis, a chain-based evidence storage mechanism is introduced to sequentially associate each early warning record to form an immutable data chain. Through chain-based evidence storage digests, key data can be efficiently extracted and transmitted. This ensures data integrity and security while reducing the computational and communication overhead of subsequent consistency verification and node collaborative processing, thereby improving the data reliability and processing efficiency of the system in complex operating environments.

[0036] Optionally, communication quality assessment is performed based on inter-node connectivity. The specific process for determining the primary backup node and candidate node list based on the communication quality assessment and establishing data backup relationships is as follows: For any two edge nodes, at least N connectivity probe messages are sent within a fixed measurement period, and the round-trip delay for each probe is recorded. The minimum round-trip delay is taken as the actual communication delay between the node pairs, and the minimum actual communication delay among all node pairs is taken as the globally optimal delay. The connectivity probe messages are sent through a lightweight communication protocol between edge nodes, preferably using a UDP or TCP-based heartbeat probe mechanism. The round-trip delay is calculated by recording the sending time of the probe message and the receiving time of the corresponding response message. The fixed measurement period is a preset time window, preferably 5 to 30 seconds, to ensure the real-time performance and stability of the communication quality assessment. N is a preset number of probes, preferably 3 to 10, to reduce the impact of occasional network fluctuations on the measurement results. Taking the minimum round-trip time as the actual communication delay is used to estimate the basic propagation delay of the link between nodes. This effectively eliminates the impact of instantaneous congestion, queuing delays, and short-term jitter on the measurement results, thus reflecting the upper bound of the network performance under ideal transmission conditions. Dividing the global optimal delay by the actual communication delay of each node pair yields the communication quality of the node pair. The communication quality is a dimensionless ratio used to characterize how close the current node pair is to the global optimal communication state. Its value ranges from 0 to 1, with a larger value indicating better communication performance. Optionally, the mutual backup and adaptation degree between nodes can be further evaluated based on geographical proximity, data correlation, and load complementarity. Specifically, the installation location coordinates of each edge node are obtained through the site equipment deployment records, and the Euclidean distance is calculated. The Euclidean distance is then multiplied by a distance attenuation coefficient and added to one, and the reciprocal is taken to obtain the geographical proximity. The attenuation coefficient can be selected according to the node spacing within the site, with a value range of (0,1). This ensures that the geographical proximity is closer to one when the nodes are closer, and gradually decreases when the nodes are farther apart, thereby reflecting the proximity of two edge nodes in the spatial distribution of the site.By utilizing the charging pile controller, the site operation management platform, and local early warning records of edge nodes, the charging start time, charging end time, early warning type, and early warning occurrence time of each edge node's corresponding charging pile location are obtained. Within a preset statistical window, the overlap duration of charging periods between two edge nodes is calculated. This overlap duration is then divided by the larger of the total charging duration of each edge node to obtain the charging period overlap. Simultaneously, the number of identical early warning types generated by two edge nodes is counted and divided by the larger of the total number of early warning types for both edge nodes to obtain the early warning type similarity. The charging period overlap and early warning type similarity are then weighted proportionally and summed to obtain the data correlation, with a preferred weighting of 0.5. This ensures that the more frequently two edge nodes engage in charging activities and generate the most identical early warnings within the same time period, the higher the data correlation, thus characterizing the similarity between the two edge nodes in their operational status and early warning behavior. By using the edge node operating system resource monitoring interface, local evidence storage cache, and mutual backup synchronization task queue, the CPU utilization rate, memory utilization rate, local storage utilization rate, and network bandwidth utilization rate of each edge node are obtained, and the average of these resource utilization rates is calculated to obtain the node resource utilization rate. At the same time, the CPU idle rate, memory remaining rate, local storage remaining capacity ratio, and network bandwidth remaining ratio are obtained, and the average of these remaining resource ratios is calculated to obtain the node remaining resource ratio. Then, the remaining resource ratio of the first edge node is multiplied by the resource utilization rate of the second edge node, and the remaining resource ratio of the second edge node is multiplied by the resource utilization rate of the first edge node. The sum of the two products is divided by the sum of the remaining resource ratios of the two edge nodes to obtain the load complementarity. When one edge node has a high load and the other edge node has sufficient remaining resources, the load complementarity is high, so as to ensure that when the primary mutual backup node is unavailable, the backup node has sufficient resources to handle mutual backup data and continuous evidence storage tasks. After normalizing geographic proximity, data correlation, load complementarity, and the aforementioned communication quality to the range of 0 to 1, a weighted sum is performed to obtain the inter-node backup affinity score. The sum of the weights of each item is 1. The data correlation score has the highest weight, preferably 0.30, because the similarity of charging time and alarm type directly affects the practical value of backup data. Geographic proximity and load complementarity are next, each preferably 0.25, as they jointly ensure data synchronization efficiency and resource balance. The communication quality score has a relatively low weight, preferably 0.20, because bandwidth is usually not a bottleneck in a site LAN environment. Maintenance personnel can adaptively adjust these proportions according to the site network topology and load characteristics.For each edge node, the communication quality between it and all other edge nodes is calculated. Optionally, the mutual backup affinity score between the current edge node and all other edge nodes can also be calculated. The edge node with the highest communication quality or mutual backup affinity score is selected as the primary mutual backup node for the current edge node, and these nodes are sorted from highest to lowest according to their communication quality or mutual backup affinity scores. The top M edge nodes ranked after the primary mutual backup node are selected as a candidate node list; where M is the preset number of candidate nodes, preferably 2 to 5, to provide redundant takeover capability when the primary mutual backup node is unavailable. When the number of edge nodes in the site is small, the node load changes little, or the network structure is relatively stable, the primary mutual backup node and candidate node list are directly selected based on communication quality to reduce the computational complexity of mutual backup pairing and improve the pairing response speed. When the number of edge nodes in the site is large, the node load fluctuates greatly, or there are cross-regional deployments or differences in services across different charging areas, the primary mutual backup node and candidate node list are selected based on mutual backup affinity score to avoid problems such as high-load nodes being mutually bound, insufficient service relevance, or insufficient takeover capability of candidate nodes caused by selecting mutual backup nodes solely based on communication quality. When the mutual backup affinity scores are the same or the difference is less than a preset difference threshold, the edge node with higher communication quality is preferentially selected as the primary mutual backup node to ensure the transmission stability and timeliness during the mutual backup data synchronization process. A data mutual backup relationship is established between the current edge node and the corresponding primary mutual backup node. Under this relationship, after generating each standardized warning record, the current edge node sends the standardized warning record, its corresponding warning record number, and the chained evidence value to the primary mutual backup node. The sending process is implemented through a reliable transmission mechanism, preferably using a transmission method with acknowledgment to ensure the integrity and order of data transmission. After receiving the standardized warning records, the primary mutual backup node stores them in the order of the warning record number and calculates the corresponding chained evidence value for each received standardized warning record, performing consistency verification. Consistency verification is achieved by comparing the received chained evidence value with the locally calculated result. When they match, the data is determined not to have been tampered with; when inconsistency exists, an abnormal record marker is triggered, and retransmission is performed to ensure the integrity and reliability of the data chain. Furthermore, after the data backup relationship is established, the current edge node can also send chain-based evidence digests to the primary backup node according to a preset digest exchange cycle. The chain-based evidence digest includes the current edge node's most recent preset number of chain-based evidence value sequences, the corresponding warning record number range, the node's unique identifier, and the digest generation time. Upon receiving the chain-based evidence digest, the primary backup node compares it with the data in the corresponding edge node's chain-based evidence value sequence with the data in the same warning record number range stored locally. When the chain-based evidence values ​​corresponding to the same warning record number are consistent, the neighbor digest verification is deemed successful. When there are inconsistent chain-based evidence values ​​or missing number ranges, the corresponding warning record range is marked as an abnormal range, and the retransmission of the standardized warning record, warning record number, and chain-based evidence value within that abnormal range is triggered.Through the aforementioned periodic summary exchange, neighbor cross-verification between edge nodes is achieved without repeatedly transmitting all original early warning records.

[0037] In this implementation plan, a communication quality assessment mechanism based on connectivity detection is used to achieve quantitative analysis of communication performance between nodes and adaptive selection of optimal backup nodes. Combined with a candidate node mechanism, the system's redundancy takeover capability in the event of node anomalies is improved. At the same time, by establishing data backup relationships and a chain-based evidence consistency verification mechanism, the integrity and reliability of early warning data transmission are guaranteed, thereby improving the system's stability, reliability, and data security in a dynamic network environment.

[0038] Optionally, the specific process of executing the backup takeover and takeover request sending when the primary node is unreachable is as follows: For each edge node, heartbeat messages are sent to the primary and backup nodes according to a fixed heartbeat detection period, and the heartbeat sending time and response receiving time are recorded. The heartbeat messages are periodically sent lightweight probe messages, preferably encapsulated using a unified communication protocol, and include at least the node's unique identifier, current timestamp, and status identifier information. The fixed heartbeat detection period is a pre-set time interval, preferably 1 to 5 seconds, used to reduce communication overhead while ensuring real-time detection. The heartbeat sending time and response receiving time are recorded based on the same time base to ensure the accuracy of round-trip delay calculation. When no heartbeat response is received from the corresponding primary and backup node within K consecutive heartbeat detection periods, the primary and backup node is determined to be in an unreachable state, where K is a preset number of consecutive determinations, preferably 3 to 5, used to avoid misjudgments due to instantaneous network jitter. When the primary backup node is unreachable, the edge node with the highest communication quality is selected from the candidate node list as the new primary backup node. A takeover request message is then sent to the new primary backup node. The candidate node list is derived from the communication quality ranking, with the edge node with the highest communication quality serving as the priority takeover node to ensure communication efficiency and stability during the takeover process. The takeover request message includes the current edge node's maximum warning record sequence number, chained evidence digest, and unique node identifier. The maximum warning record sequence number identifies the latest position of the data generated by the current edge node, the chained evidence digest describes the status information of the most recent chained evidence value sequence, and the unique node identifier distinguishes different data source nodes. The takeover request message is sent through a reliable transmission mechanism, preferably using a communication method with acknowledgment to ensure that the takeover information is fully received by the new primary backup node, providing a foundation for subsequent data alignment and incremental recovery. Furthermore, the chain-based evidence digest carried in the takeover request not only describes the current state of the data chain tail, but also serves as a key basis for subsequent alignment matching and determination of common data boundaries. This enables the new primary and backup nodes to quickly locate the data consistency interval between the two parties based on the digest information, thereby reducing the overhead of full comparison and avoiding duplicate data transmission. By transmitting the chain-based evidence digest in advance during the takeover phase, the fault detection and data recovery processes are connected, so that the takeover action not only completes the node role switch, but also provides direct input for subsequent boundary alignment and data recovery, thereby enhancing the continuity and recovery efficiency of the overall process.

[0039] In this implementation scheme, a master node reachability determination mechanism based on heartbeat detection is used to realize real-time perception of the status of master and backup nodes. When a node becomes unreachable, the adaptive takeover and takeover request triggering mechanism of the backup node is used to quickly complete the master node switchover and data takeover. At the same time, combined with the synchronous transmission of key status information, a reliable basis is provided for subsequent data alignment and recovery, thereby improving the system's continuous operation capability and self-healing recovery capability in the event of network anomalies or node failures.

[0040] Optionally, for the node being taken over, the specific process of performing alignment matching based on the chain-based evidence digest is as follows: The new primary backup node receives and parses the takeover request message. The parsing process is used to extract the maximum warning record number, the chain-based evidence digest, and the node's unique identifier from the takeover request message, and to locate the data storage partition of the corresponding edge node based on the node's unique identifier. Simultaneously, the new primary backup node reads the standardized warning record sequence of the corresponding edge node that it has already stored, determines the local maximum warning record number, and calculates the chain-based evidence value one by one based on the stored standardized warning records, forming a local chain-based evidence value sequence. The calculation method for the chain-based evidence value is consistent with that of the edge node side to ensure the reproducibility and consistency of the chain structure. The local maximum warning record number is determined by traversing the largest warning record number among the stored records. Records of the same length as the chain-based evidence digest are traced back from the local maximum warning record number, and the corresponding chain-based evidence value sequence is extracted to form a local chain-based evidence digest. The length of the chain-based evidence digest is a preset fixed value, preferably 10 to 30 records, to achieve a balance between matching accuracy and computational overhead. The chained evidence digest of the current edge node is aligned and matched with the local chained evidence digest. Starting from the end value of the local chained evidence digest, the matching position closest to the end of the chained evidence digest of the current edge node with the same chained evidence value is found as the alignment position. This ensures that the most recently generated data is matched first, reducing the interference of historical duplicate fragments on the matching results. When multiple identical chained evidence values ​​exist, only the one closest to the end is selected as the unique alignment benchmark. Then, starting from the alignment position, each item is compared backwards. When the corresponding chained evidence values ​​are equal, the count is incremented by one; when they are unequal, the comparison stops. The longest consecutive identical suffix length is obtained. The longest consecutive identical suffix length is used to characterize the degree of continuity and consistency of the chained data of two nodes at the end, which is an important basis for subsequent boundary alignment and data recovery. Since the chained evidence values ​​are generated using SHA-256, their collision probability is lower than the preset security threshold. Therefore, identical chained evidence values ​​can be considered to correspond to the same original record sequence. The longest consecutive identical suffix length is the consecutive matching length obtained by comparing items backward from the alignment position. The comparison object is a chain-like sequence of stored evidence values, and each chain-like stored evidence value is bound to its corresponding warning record number. During the comparison process, the consistency of the chain-like stored evidence values ​​and the continuous incrementing of the warning record number are ensured, thereby avoiding mismatches caused by misaligned numbers. By identifying the longest consecutive identical suffix, the common consistent interval at the end of the chain-like data of two nodes can be determined. The data segment corresponding to the consistent interval is the data range that both parties have confirmed, thereby accurately locating the common data boundary and effectively avoiding data duplication synchronization or data breakpoint problems.The round-trip time (RTT) of historical connectivity probe messages between the current edge node and the new primary / backup node is statistically analyzed, and the mean and standard deviation of the RTT are calculated. The RTT is determined based on connectivity probe data within a fixed historical time window prior to the takeover, calculated by the difference between the sending and receiving times of the corresponding probe messages, ensuring the real-time nature and validity of the statistical results. The time window is preferably 10 to 60 seconds, and the number of samples included in the statistics is no less than the preset minimum sample size, preferably no less than 5 times, to ensure the stability and representativeness of the calculated mean and standard deviation of the RTT. When the number of valid samples is insufficient, the statistical window can be extended to supplement the data, avoiding excessive fluctuations in the statistical results due to insufficient samples. The mean RTT is used to characterize the communication latency level, and the standard deviation of the RTT is used to characterize the degree of communication fluctuation, thus providing a basis for communication stability assessment in subsequent alignment reliability evaluation.

[0041] In this implementation scheme, an alignment and matching mechanism based on chain-based evidence digests is used to achieve precise location and quantitative characterization of data chain tail consistency between different nodes. By combining round-trip delay statistics to introduce communication stability factors, alignment judgment is supported from two dimensions: data consistency and communication status. This provides a reliable basis for subsequent data boundary determination and incremental recovery, and improves the data alignment accuracy and recovery efficiency of the system in node switching scenarios.

[0042] Optionally, the specific process for determining the consistency of the chain tail and the degree of boundary offset to form the boundary reliability assessment result is as follows: divide twice the length of the longest consecutive identical suffix by the sum of the length of the chain evidence digest sent by the current edge node and the length of the local chain evidence digest to obtain the chain tail overlap ratio; wherein, the length of the chain evidence digest is a preset fixed value, preferably 10 to 30, the length of the longest consecutive identical suffix is ​​obtained by the aforementioned alignment matching process, and is used to characterize the continuity and consistency of the chain tail data of the two nodes, and the chain tail overlap ratio is a dimensionless ratio, the value range of which is 0 to 1, the larger the value, the higher the consistency of the chain tail, which can reflect the direct matching relationship at the data content level. Calculate the absolute difference between the maximum early warning record number of the current edge node and the maximum early warning record number locally, and take the minimum of the two. Divide the absolute difference by the sum of the minimum and the smallest positive number to obtain the relative boundary offset. Negatively calculate the relative boundary offset and then perform a natural exponential operation to obtain the boundary offset suppression term. The smallest positive number is a preset non-zero constant, preferably within a certain range. arrive The term is used to avoid zero denominators and improve computational stability. The boundary offset suppression term ranges from 0 to 1 and decreases monotonically as the difference in warning record sequence numbers increases, reflecting the impact of data progress offset between nodes on alignment reliability. By introducing an exponential decay function, the impact of sequence number offset on alignment reliability is nonlinearly compressed, resulting in a weaker impact when the offset is small and a significantly stronger impact when the offset is large, thus conforming to the variation law of error accumulation during actual data synchronization. The delay dispersion ratio is obtained by dividing the round-trip delay standard deviation by the sum of the round-trip delay mean and the smallest positive number. The delay dispersion ratio is then increased by one and the reciprocal is taken to obtain the delay stability term. The round-trip delay standard deviation and the round-trip delay mean are statistically obtained based on connectivity detection data before takeover. The delay stability term is a dimensionless index, ranging from 0 to 1. A larger value indicates a more stable communication state, used to characterize the impact of communication link stability on data alignment results, avoiding misjudgments under conditions of large communication fluctuations. The boundary alignment value is obtained by multiplying the chain tail overlap ratio term, the boundary offset suppression term, and the delay stability term. Boundary alignment value is a comprehensive evaluation index that integrates three dimensions: chain tail consistency, boundary offset, and communication stability. Its value ranges from 0 to 1, with a larger value indicating higher reliability of data alignment between nodes. The three indicators are fused through multiplication, so that a significant decrease in any key factor can suppress the overall alignment result. This ensures that the boundary alignment value only reaches a large value when chain tail consistency is high, sequence number offset is small, and communication is stable. This allows for an accurate representation of the reliability of data alignment between nodes and can serve as a basis for determining the starting position of subsequent data recovery and selecting synchronization strategies.

[0043] The specific formula for the boundary alignment value is as follows:

[0044] ;

[0045] In the formula, This represents the boundary alignment value, used to comprehensively evaluate the data alignment reliability between the current edge node and the new primary backup node. By integrating chain tail consistency, boundary offset degree and communication stability, it determines the credibility of the data synchronization boundary, thereby providing a basis for determining the subsequent recovery interval. This indicates the length of the longest consecutive identical suffix, representing the degree of consistency between the tail data of two nodes; This indicates the length of the chained evidence digest sent by the current edge node, representing the range of data for which the current node participates in alignment; Indicates the length of the local chained evidence digest, representing the data range in which the takeover node participates in alignment; This represents the maximum warning record number of the current edge node, indicating the latest data position of the current node; This indicates the sequence number of the highest local warning record, representing the latest data position of the takeover node; It represents the standard deviation of round-trip delay, characterizing the degree of communication fluctuation; This represents the average round-trip time, characterizing the level of communication latency. This represents a very small positive number, used to prevent the denominator from being zero and to stabilize the calculation process; the preferred value is [value missing]. arrive .

[0046] In this embodiment, Table 1 is a boundary alignment value data table. The smallest positive number is... The length of the chained evidence digest sent by the current edge node and the length of the local chained evidence digest are both 20. The table details the longest consecutive identical suffix length, the maximum warning record number of the current edge node, the maximum local warning record number, the round-trip delay standard deviation, the mean round-trip delay, and the boundary alignment value for five node pairs. Specifically, for node pair AD, the longest consecutive identical suffix length is 7, the maximum warning record number of the current edge node is 120, the maximum local warning record number is 92, the round-trip delay standard deviation is 3.6, the mean round-trip delay is 24, and the boundary alignment value is 0.22449; for node pair BF, the longest consecutive identical suffix length is 3, the maximum warning record number of the current edge node is 118, the maximum local warning record number is 60, the round-trip delay standard deviation is 4.8, the mean round-trip delay is 29, and the boundary alignment value is 0.04895; for node pair CA, the longest consecutive identical suffix length is 16, the maximum warning record number of the current edge node is 110, and the local... The maximum warning record number is 108, the standard deviation of round-trip delay is 1.5, the mean of round-trip delay is 21, and the boundary alignment value is 0.73297; the longest consecutive identical suffix length corresponding to the node pair DC is 9, the maximum warning record number of the current edge node is 102, the local maximum warning record number is 95, the standard deviation of round-trip delay is 2.8, the mean of round-trip delay is 22, and the boundary alignment value is 0.37084; the longest consecutive identical suffix length corresponding to the node pair EB is 11, the maximum warning record number of the current edge node is 90, the local maximum warning record number is 75, the standard deviation of round-trip delay is 2.2, the mean of round-trip delay is 23, and the boundary alignment value is 0.41099.

[0047] surface Boundary Alignment Value Data Table

[0048]

[0049] like Figure 3The figure shows a three-dimensional evaluation diagram of boundary alignment reliability based on multi-factor fusion. The diagram visualizes the data alignment reliability of different node pairs in three-dimensional space. The horizontal axis represents the chain tail overlap ratio, characterizing the consistency of the tail of the chain-like evidence sequence; the vertical axis represents the boundary offset suppression, reflecting the degree of suppression of the warning record sequence number offset; and the vertical axis represents the latency stability, characterizing the communication latency fluctuation between nodes. Each scatter point in the diagram corresponds to a different node pair, and the labeled value is the corresponding boundary alignment value. The scatter point size increases with the boundary alignment value. (See Table 1 and...) Figure 3 It can be seen that node pair CA is located in the high-value region of the three-dimensional space, with high chain tail consistency, small sequence number offset, and stable communication, corresponding to the largest boundary alignment value, indicating the highest data synchronization reliability. Node pair BF is located in the low-value region, with poor chain tail consistency, large sequence number offset, and significant latency fluctuation, corresponding to the smallest boundary alignment value, indicating a significant synchronization risk. Other node pairs such as EB and DC are in the middle region, reflecting moderate consistency and a recoverable state. AD shows partial consistency but with historical synchronization lag. It can be seen that the boundary alignment value increases with the increase of the chain tail overlap ratio, decreases with the increase of the sequence number offset, and is modulated by latency stability. These three factors jointly determine the reliability of data synchronization between nodes, thus effectively integrating multiple factors to achieve a comprehensive assessment of boundary alignment reliability.

[0050] In this implementation plan, by constructing a chain tail overlap ratio term, a boundary offset suppression term, and a time delay stability term, a comprehensive quantitative assessment of the reliability of data alignment between nodes from three dimensions—data content consistency, data progress offset, and communication stability—is achieved. Furthermore, dimensionless normalization and exponential suppression mechanisms enhance the model's sensitivity and stability to offsets and fluctuations. Simultaneously, a multiplicative fusion approach is employed to effectively suppress the overall assessment results even if any key factor anomaly occurs, thereby improving the accuracy and robustness of alignment determination and providing a reliable basis for determining the starting position of subsequent data recovery and optimizing synchronization strategies.

[0051] Optionally, based on the boundary reliability assessment results, the specific process of identifying the data recovery start position and recovery interval to achieve data completion and incremental recovery is as follows: Multiply the difference between the boundary alignment value and the boundary alignment value by the length of the longest consecutive identical suffix, and round up to obtain the boundary rollback count. The boundary alignment value is the aforementioned comprehensive assessment result, ranging from 0 to 1. The boundary rollback count is used to adaptively adjust the matching boundary according to the alignment reliability. When the boundary alignment value is small, the rollback count increases accordingly to avoid mistakenly including inconsistent data in the alignment range. Furthermore, the boundary rollback mechanism actively expands the rollback range when the boundary alignment reliability is low, thereby covering potential inconsistent areas and eliminating hidden breakpoints caused by mismatches. When the boundary alignment value is high, the rollback count decreases, making the common data boundary closer to the true consistent position, thereby reducing the amount of duplicate data synchronization and improving synchronization efficiency while ensuring data integrity. This mechanism achieves a balance between "misjudged boundaries" and "duplicated synchronization" through the linkage adjustment of alignment reliability and rollback magnitude. The common data boundary number is obtained by subtracting the longest consecutive identical suffix length from the maximum warning record number of the current edge node, then adding one, and finally subtracting the number of boundary backtracking entries. The subtraction of the longest consecutive identical suffix length from the maximum warning record number of the current edge node, plus one, is used to convert the "length information" of the consecutive consistent data at the chain tail into the corresponding "sequence start position," thus determining the starting point of the consistent interval at the chain tail. Since the longest consecutive identical suffix length represents the number of consecutively matched data entries and does not contain specific sequence position information, obtaining the first record number of the consistent data interval using the method of "maximum sequence number - length + 1" conforms to the general rules for determining continuous sequence intervals. Furthermore, the number of boundary backtracking entries is subtracted to extend and adjust the boundary position forward when alignment reliability is insufficient, thereby covering potential inconsistencies and avoiding boundary misjudgments due to mismatches or accidental consistency at the chain tail. Through the above processing, the final common data boundary number reflects the position of the consistent interval at the chain tail and can be adaptively corrected based on alignment reliability, thus reducing the risk of duplicate synchronization while ensuring data integrity. The common data boundary number is used to identify the boundary position of data consistency between two nodes. Data before the common data boundary number is considered as aligned data, while data after the common data boundary number is considered as data to be verified and synchronized. The common data boundary number is limited to a valid range, with a value of at least 1 and not greater than the minimum value between the current edge node and the local maximum warning record number, to ensure that the boundary position is legal and does not exceed the data range of both nodes.When the local maximum warning record number is less than the common data boundary number, it is determined that there is a data gap before the common data boundary in the new primary backup node. The interval between the local maximum warning record number plus one and the common data boundary number is read as the missing data filling interval. The current edge node sends the standardized warning record, the corresponding warning record number, and the chained evidence value within the missing data filling interval to the new primary backup node. The missing data filling interval is used to fill in the data not stored in the new primary backup node to ensure the continuity of the chained evidence sequence. The missing data filling interval is preferably represented in the form of a closed interval, that is, the interval is [local maximum warning record number + 1, common data boundary number], so as to clarify the inclusion relationship of the interval endpoints. After missing data completion, the interval between the common data boundary number plus one and the current edge node's maximum warning record number is read as the incremental synchronization interval. The standardized warning records, corresponding warning record numbers, and chained evidence values ​​within this interval are sent. The incremental synchronization interval is used to synchronize the latest generated warning data, enabling real-time updates of the data chain. The incremental synchronization interval is preferably represented as (common data boundary number, current edge node's maximum warning record number) to avoid duplicate synchronization of data at already aligned boundaries. When the local maximum warning record number is greater than or equal to the common data boundary number, it is determined that the new primary backup node has covered the data before the common data boundary. The interval between the local maximum warning record number plus one and the current edge node's maximum warning record number is read as the data synchronization interval. The current edge node sends the standardized warning records, corresponding warning record numbers, and chained evidence values ​​within this data synchronization interval to the new primary backup node. The data synchronization interval is used to perform incremental updates directly without needing to complete historical data, thereby improving synchronization efficiency. The data synchronization interval also adopts a closed interval format [local maximum warning record number + 1, current edge node maximum warning record number] to unify the interval representation rules and avoid ambiguity. After receiving the data, the new primary backup node performs storage according to the warning record number sequence, calculates the chained evidence value for each received standardized warning record, updates the local chained evidence value sequence, and returns a data synchronization completion confirmation message to the current edge node. The confirmation message is used to indicate that the synchronization process has been completed and can serve as a trigger condition for subsequent data consistency verification and status updates. Based on the above method, when there is partial matching at the end of the chain but the boundary alignment value is low, the common data boundary is moved forward by increasing the number of boundary backtracking entries, thereby covering potential inconsistencies and avoiding the omission of historical data. When the end of the chain is completely matched and the boundary alignment value is high, the common data boundary is close to the true consistent position, and only the minimum range of incremental synchronization is performed to verify that the mechanism can achieve a balance between data integrity and synchronization efficiency under different alignment reliability conditions.

[0052] In this implementation scheme, an adaptive fallback mechanism based on boundary alignment values ​​is used to accurately identify and dynamically adjust data consistency boundaries. Combined with a segmented recovery strategy of missing data completion and incremental synchronization, the data recovery efficiency is improved while ensuring the continuity and consistency of chained data. At the same time, a closed-loop control is formed through a confirmation feedback mechanism, which enhances the accuracy and reliability of data recovery in node switching scenarios.

[0053] Optionally, the specific process of determining the connection state based on the heartbeat interaction result, and performing data mutual backup, degradation adjustment and offline evidence storage control under different connection states is as follows: within a fixed statistical time window, count the number of heartbeat messages sent by the current edge node to the current primary mutual backup node and the number of successfully received responses, divide the number of successfully received responses by the number of sent heartbeat messages to obtain the connectivity C between the current edge node and the primary mutual backup node, and compare the connectivity C with multi-level connectivity thresholds C1 and C2; wherein, the fixed statistical time window is a preset time range, preferably 10 seconds to 60 seconds, which is used to兼顾 both real-time performance and statistical stability; the heartbeat message is a periodically sent lightweight detection message, which at least includes the unique node identifier and timestamp information, and the number of successful response receptions is determined by judging whether the corresponding response message is received within a preset timeout threshold; the connectivity C is a dimensionless ratio with a value range from 0 to 1, which is used to quantify the communication connection state between nodes. When it is detected that the current primary mutual backup node changes within the fixed statistical time window, the new primary mutual backup node is重新 used as the statistical object, the number of sent heartbeat messages and the number of successful response receptions within the fixed statistical time window are reset, and the connectivity C is recalculated, so as to avoid judgment deviation caused by mixed use of statistical data between different primary mutual backup nodes; meanwhile, the data mutual backup relationship is re-established based on the updated primary mutual backup node, and its connection state is continuously tracked in the subsequent heartbeat detection process. When C≥C2, it is determined that the current edge node is in a normal connection state; when C1≤C<C2, it is determined that the current edge node is in a connection degradation state; when C<C1, it is determined that the current edge node is in an offline state; wherein, C1 and C2 are preset connection thresholds, which preferably satisfy 0≤C1<C2≤1, for example, C1 can be 0.3 to 0.6, and C2 can be 0.7 to 0.9, so as to realize hierarchical judgment of different connection states. When the edge node is in a normal connection state, the generation of standardized early warning records, the data mutual backup relationship between primary mutual backup nodes and the calculation of chain evidence storage value are maintained; wherein, shortening the heartbeat detection period means reducing the original period according to a preset proportion, preferably shortening to 1 / 2 to 2 / 3 of the original period, so as to improve the frequency of state perception and detect connectivity changes in time. When the edge node is in a connection degradation state, on the premise of shortening the heartbeat detection period, the generation of standardized early warning records, the data mutual backup relationship between primary mutual backup nodes and the calculation of chain evidence storage value are maintained; wherein, in the connection degradation state, in order to avoid aggravating network congestion caused by the increase of heartbeat detection frequency, an adaptive degradation adjustment strategy can be adopted. While shortening the heartbeat detection period, throttling control is performed on the sending of early warning data, including but not limited to realizing by reducing the reporting frequency of early warning records, extending the data sending period, and uniformly sending after batch aggregation of multiple early warning records, so as to reduce link load while ensuring the perception sensitivity of the connection state; batch sending can be triggered based on a fixed quantity threshold, for example, sending is executed when the accumulated early warning records reach the preset number, so as to balance communication overhead and data real-time performance.When an edge node is offline, it suspends sending standardized warning records, warning record numbers, and chained evidence values ​​to the primary and backup nodes. It only continues to perform standardized warning record generation, chained evidence summary update, and chained evidence value calculation locally to ensure that data can still be completely recorded and the chained evidence structure can be maintained during network interruption, providing a basis for data synchronization and consistency verification after the network is restored.

[0054] In this implementation scheme, a connectivity quantification determination mechanism based on heartbeat interaction is used to achieve hierarchical identification of node communication status. Under different connectivity statuses, data backup, frequency adjustment and network outage evidence preservation control strategies are adaptively executed. While ensuring normal communication efficiency, the ability to perceive and respond to network degradation and interruption is enhanced, thereby improving data continuity, reliability and self-healing ability in complex network environments.

[0055] Optionally, the self-healing process is triggered during network recovery. The specific process for completing chain-like data consistency recovery and continuous evidence reconstruction is as follows: When the node's operating state recovers from a network outage or connectivity degradation state to a normal connectivity state, the network outage self-healing recovery process is triggered: The current primary and backup nodes of the edge nodes are determined. The current primary and backup nodes are the optimal nodes determined based on the latest communication quality assessment results or backup affinity scores to ensure communication efficiency and stability during the recovery process; when the primary and backup nodes change during the recovery process, the node with the highest communication quality or backup affinity score and is in a reachable state is selected as the new primary and backup nodes. The system sends the maximum warning record number, the corresponding chain-based evidence digest, and the node's unique identifier to the primary and backup nodes. The maximum warning record number identifies the latest position of the current edge node's data chain, the chain-based evidence digest describes the state of the most recent segment of chained data, and the node's unique identifier locates the corresponding data source. The transmission process preferably employs a reliable transmission mechanism with acknowledgment to ensure the complete delivery of critical recovery information. A boundary alignment value is calculated based on chain tail consistency, boundary offset, and communication stability, and is used to assess the reliability of data alignment between current nodes. The system also completes the determination of common data boundary numbers, the identification of missing data intervals, and the determination of data synchronization intervals. The system includes the supplementary transmission of corresponding standardized early warning records. A common data boundary number is used to identify the boundary where data from both sides is consistent. Data missing intervals are used to fill in historical data not covered by the primary and backup nodes. Data synchronization intervals are used to synchronize the latest generated data to achieve continuous data chain connection. The system also updates the standardized early warning record sequence and chain-based evidence value sequence in the primary and backup nodes. During the update process, the records are stored in the order of the early warning record sequence number, and the chain-based evidence value calculation and verification are performed synchronously to ensure the integrity and consistency of the restored chain structure. After restoration, the primary and backup nodes return a restoration completion confirmation message to the edge nodes to indicate the end of the self-healing process and trigger a status update. The boundary alignment value is calculated collaboratively by the edge nodes and the primary and backup nodes. The primary and backup nodes perform alignment matching based on the received chain-based evidence digest from the corresponding edge node and their local chain-based evidence digest, and complete calculations related to chain tail consistency and boundary offset. Simultaneously, the edge nodes provide communication latency statistics as auxiliary input, clarifying the calculation inputs, outputs, and responsible parties to avoid duplicate and conflicting calculations on different nodes. After recovery, the edge nodes also send the chained evidence value sequence generated during the network outage, the corresponding warning record number range, the node's unique identifier, and the locally generated time series to the cloud platform. Among them, the chained evidence value sequence is used to represent the locally chained evidence results generated continuously during the network outage, the warning record number range is used to identify the data range covered by the corresponding chained evidence value sequence, and the locally generated time series is used to represent the generation order of each chained evidence value during the network outage.After receiving the aforementioned data, the cloud platform reads the last cloud anchor point saved by the current edge node before the network outage and the latest cloud anchor point after network recovery. It then performs continuity checks on the chained evidence value sequence re-transmitted during the network outage, according to the warning record sequence number. The continuity check is achieved by determining whether the first chained evidence value re-transmitted during the network outage forms a continuous correspondence with the chained evidence value corresponding to the last cloud anchor point before the outage, and whether the last chained evidence value re-transmitted during the network outage forms a continuous correspondence with the chained evidence value corresponding to the latest cloud anchor point after network recovery. The cloud platform also performs time sequence checks on the chained evidence value sequence re-transmitted during the network outage according to the locally generated time sequence to determine whether the generation time of each chained evidence value during the network outage satisfies a time-increasing relationship. When the chained evidence value sequence re-transmitted during the network outage satisfies both the continuity check and the time sequence check, the continuous evidence storage during the network outage is deemed valid, and the corresponding chained evidence record in the cloud platform is updated, thus completing the cloud-based backtracking verification after network recovery.

[0056] In this implementation scheme, by triggering a self-healing process when the network is restored, and combining a chain-style evidence digest and boundary alignment evaluation mechanism, the accurate identification of data consistency boundaries and the orderly retransmission of missing data are achieved. While ensuring the continuity of the chain structure, the collaborative restoration of historical data and incremental data is completed. A closed-loop control is formed through confirmation feedback, thereby improving the data reconstruction efficiency and consistency reliability in network outage recovery scenarios.

[0057] like Figure 2 As shown, the second aspect of the present invention provides a self-healing system for site early warning and evidence storage based on peer-to-peer edge redundancy, applied to the aforementioned self-healing method for site early warning and evidence storage based on peer-to-peer edge redundancy, comprising: a site sensing data processing module, used to collect multi-source operational data from edge nodes in real time, perform time-series unification and validity screening on the multi-source operational data, form standardized early warning records based on early warning judgment, construct a continuous chain-like evidence storage sequence, and extract chain-like evidence storage summaries; a peer-to-peer edge redundancy takeover module, used to perform communication quality assessment based on inter-node connectivity, determine a list of primary redundancy nodes and candidate nodes based on the communication quality assessment, establish a data redundancy relationship, and on the primary node... When unreachable, a backup takeover and takeover request are sent. The continuous evidence preservation boundary recovery module is used to perform alignment matching based on the chain evidence digest for the takenover node, determine the consistency degree of the chain tail and the degree of boundary offset, form a boundary reliability assessment result, and identify the data recovery start position and recovery interval based on the boundary reliability assessment result to realize data completion and incremental recovery. The evidence preservation state switching self-healing module is used to determine the connectivity state based on the heartbeat interaction result, perform data backup, degradation adjustment and network disconnection evidence preservation control under different connectivity states, and trigger the self-healing process when the network is restored to complete the chain data consistency recovery and continuous evidence preservation reconstruction.

[0058] like Figure 4 The diagram shows the process flow of network outage self-healing collaborative recovery based on chained evidence storage and boundary alignment. It illustrates the overall process of a site early warning evidence storage self-healing method based on peer-to-peer edge redundancy. First, multi-source operational data from edge nodes is collected in real time. The data undergoes time-series unification and validity screening to generate standardized early warning records, and a chained evidence storage sequence and chained evidence storage summary are constructed. Then, based on the inter-node communication quality assessment results, primary and backup nodes are determined, and a data redundancy relationship is established. Simultaneously, the connectivity status of the primary and backup nodes is continuously monitored through heartbeat detection. When the primary and backup nodes are reachable, normal data redundancy and chained evidence storage calculations are maintained. When the primary and backup nodes are detected as unreachable, the backup node performs takeover and sends a takeover request. Subsequently, alignment matching is performed based on the chained evidence storage summary to calculate the boundary alignment degree and determine the data consistency boundary. Further determination of the missing data completion interval and incremental synchronization interval completes data recovery and chained sequence updates. During operation, adaptive adjustments are made based on the connectivity status determination results. When in a connectivity degradation state, the detection capability is enhanced by shortening the heartbeat cycle. When in a network outage state, remote synchronization is suspended, and only local chain-based evidence is retained to ensure data continuity. When the network is restored, the network outage self-healing process is triggered to complete data replenishment, incremental synchronization, and chain consistency reconstruction, ultimately restoring normal operation and forming a closed-loop control process.

[0059] In this implementation plan, by constructing modules such as data acquisition and processing, mutual backup pairing takeover, boundary alignment recovery, and network outage self-healing control, a closed-loop management of the entire process from data generation, node collaboration, alignment recovery to state regulation is realized. On this basis, through a boundary alignment mechanism that combines chain-based evidence storage and multi-factor fusion, combined with heartbeat-driven connectivity state adaptive control and backup takeover strategy, data continuity is maintained and consistency is restored in network anomaly and node switching scenarios, thereby improving the system's reliability, robustness, and self-healing capabilities.

[0060] The above embodiments can be implemented, in whole or in part, by software, hardware (such as circuits), firmware, or any other combination thereof. When implemented using software, the above embodiments can be implemented, in whole or in part, as a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, all or part of the processes or functions described in the embodiments of the present invention are generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that includes one or more sets of available media. The available medium can be a magnetic medium (e.g., floppy disk, hard disk, magnetic tape), an optical medium (e.g., DVD), or a semiconductor medium. A semiconductor medium can be a solid-state drive.

[0061] It should be understood that the term "and / or" in this article is merely a description of the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A alone, A and B simultaneously, or B alone. A and B can be singular or plural. Additionally, the character " / " in this article generally indicates an "or" relationship between the preceding and following related objects, but it can also represent an "and / or" relationship. Please refer to the context for a more accurate understanding.

[0062] In this invention, "at least one" means one or more, and "more than one" means two or more. "At least one of the following" or similar expressions refer to any combination of these items, including any combination of a single item or a plurality of items. For example, at least one of a, b, or c can represent: a, b, c, ab, ac, bc, or abc, where a, b, and c can be a single item or multiple items.

[0063] It should be understood that, in various embodiments of the present invention, the order of the above-mentioned process numbers does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present invention.

[0064] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementations should not be considered beyond the scope of this invention.

[0065] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working processes of the devices, apparatuses, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0066] In the several embodiments provided by this invention, it should be understood that the disclosed devices, apparatuses, and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another device, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between devices or units may be electrical, mechanical, or other forms.

[0067] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0068] In addition, the functional units in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit.

[0069] If the aforementioned functions are implemented as software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this invention, or the part that contributes to the prior art, or a portion of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory, random access memory, magnetic disks, or optical disks.

[0070] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A self-healing method for early warning and evidence storage at stations based on peer-to-peer edge redundancy, characterized in that: The method includes: S1 collects multi-source operation data from edge nodes in real time, performs time-series unification and validity screening on the multi-source operation data, forms standardized early warning records based on early warning judgment, constructs a continuous chain of evidence storage sequence, and extracts chain evidence storage summary. S2, based on the connectivity between nodes, performs communication quality assessment, determines the list of primary backup nodes and candidate nodes according to the communication quality assessment, establishes data backup relationship, and executes candidate takeover and takeover request sending when the primary node is unreachable; S3, for the nodes under takeover, performs alignment matching based on the chain-stored digest, determines the consistency of the chain tail and the degree of boundary offset, forms the boundary reliability assessment result, and identifies the data recovery start position and recovery interval based on the boundary reliability assessment result, so as to realize data completion and incremental recovery; S4 determines the connectivity status based on the heartbeat interaction results, performs data backup, degradation adjustment and network outage evidence preservation control under different connectivity statuses, and triggers a self-healing process when the network recovers to complete chain-like data consistency recovery and continuous evidence preservation reconstruction.

2. The self-healing method for site early warning and evidence storage based on peer-to-peer edge redundancy as described in claim 1, characterized in that, The process of collecting multi-source operational data from edge nodes in real time, performing time-series unification and validity screening on the multi-source operational data, forming standardized early warning records based on early warning judgment, constructing a continuous chain of evidence storage sequences, and extracting chain evidence storage digests is as follows: Edge nodes collect raw operating data in real time based on the sensor modules of the corresponding charging pile locations. The raw operating data includes: voltage, current, temperature, insulation resistance, node reception time, and node unique identifier. The node reception time is used as a unified time reference. Time-unified processing is performed on each piece of raw operating data collected, and it is recorded according to a fixed sampling period. For any edge node, if there are two or more raw operating data records within the fixed sampling period, only the record with the earliest node reception time is retained. For the remaining raw operating data, it is determined whether the raw operating data is within the physical valid range. If it does not meet the requirement, the corresponding raw operating data record is deleted. For each original operational data record, when the temperature exceeds the temperature warning threshold or the current exceeds the current warning threshold, it is determined to be a warning record; for warning records, they are sorted in ascending order according to the node reception time within the same edge node, and a continuously increasing warning record number is generated; the warning record number and the corresponding original operational data are combined to form a standardized warning record; Following the sequence number of the early warning records within the same edge node, the current standardized early warning record is concatenated with the previous chain-stored evidence value and then input into a hash function to calculate the current chain-stored evidence value; a chain-stored evidence value sequence of a fixed length is traced back from the current maximum early warning record number to serve as the chain-stored evidence digest.

3. The self-healing method for site early warning and evidence storage based on peer-to-peer edge redundancy as described in claim 1, characterized in that, The specific process of conducting communication quality assessment based on inter-node connectivity, determining the primary backup node and candidate node list based on the communication quality assessment, and establishing data backup relationships is as follows: For any two edge nodes, send no less than N connectivity probe messages within a fixed measurement period and record the round-trip delay of each probe. Take the minimum round-trip delay as the actual communication delay between the node pairs and take the minimum actual communication delay among all node pairs as the global optimal delay. The communication quality of a node pair can be obtained by dividing the global optimal delay by the actual communication delay of the node pair. For each edge node, the communication quality with all other edge nodes is calculated. The edge node with the highest communication quality is selected as the primary backup node for the current edge node. The edge nodes are then sorted from highest to lowest communication quality, and the top M edge nodes ranked after the primary backup node are selected as candidate nodes. A data backup relationship is established between the current edge node and its corresponding primary backup node. Under this data backup relationship, after the current edge node generates each standardized warning record, it sends the standardized warning record, its corresponding warning record number, and the chained evidence value to the primary backup node. Upon receiving the standardized warning record, the primary backup node stores it in the order of the warning record number and calculates the corresponding chained evidence value for each received standardized warning record, performing consistency verification.

4. The self-healing method for site early warning and evidence storage based on peer-to-peer edge redundancy as described in claim 1, characterized in that, The specific process of executing the alternate takeover and sending the takeover request when the master node is unreachable is as follows: For each edge node, heartbeat messages are sent to the primary backup node according to a fixed heartbeat detection period, and the heartbeat sending time and response receiving time are recorded. If no heartbeat response is received from the corresponding primary backup node within K consecutive heartbeat detection periods, the primary backup node is determined to be in an unreachable state. When the primary backup node is in an unreachable state, the edge node with the highest communication quality is selected from the candidate node list as the new primary backup node, and a takeover request message is sent to the new primary backup node. The takeover request message includes the maximum warning record sequence number of the current edge node, the chained evidence digest, and the node's unique identifier.

5. The self-healing method for site early warning and evidence storage based on peer-to-peer edge redundancy as described in claim 1, characterized in that, The specific process of performing alignment matching based on the chained evidence digest for the nodes under takeover is as follows: The new primary backup node receives and parses the takeover request message. At the same time, the new primary backup node reads the standardized early warning record sequence of the corresponding edge node that it has stored, determines the local maximum early warning record number, and calculates the chained evidence value one by one based on the stored standardized early warning records to form a local chained evidence value sequence. It then backtracks from the local maximum early warning record number to the record with the same length as the chained evidence digest, extracts the corresponding chained evidence value sequence, and forms a local chained evidence digest. Align and match the chained evidence digest of the current edge node with the local chained evidence digest. Starting from the end value of the local chained evidence digest, find the matching position in the chained evidence digest of the current edge node with the same chained evidence value that is closest to the end of the chained evidence digest as the alignment position. Start from the alignment position and compare item by item backward. When the corresponding chained evidence values ​​are equal, increment the count by one. Stop the comparison when they are not equal. Obtain the length of the longest consecutive identical suffix. Calculate the round-trip time of the historical connectivity probe messages between the current edge node and the new primary backup node, and calculate the mean and standard deviation of the round-trip time.

6. The self-healing method for site early warning and evidence storage based on peer-to-peer edge redundancy as described in claim 1, characterized in that, The specific process for determining the consistency of the chain tail and the degree of boundary offset to form the boundary reliability assessment result is as follows: Divide twice the length of the longest consecutive identical suffix by the sum of the length of the chained evidence digest sent by the current edge node and the length of the local chained evidence digest to obtain the chain tail overlap ratio. Calculate the absolute difference between the maximum early warning record number of the current edge node and the maximum early warning record number of the local node, and take the minimum value between the maximum early warning record number of the current edge node and the maximum early warning record number of the local node. Divide the absolute difference by the sum of the minimum value and the smallest positive number to obtain the relative boundary offset. After taking the negative value of the relative boundary offset, perform natural exponentiation to obtain the boundary offset suppression term. Divide the standard deviation of the round-trip delay by the sum of the mean of the round-trip delay and the smallest positive number to obtain the delay dispersion ratio. Add one to the obtained delay dispersion ratio and take the reciprocal to obtain the delay stability term. Multiply the chain tail coincidence ratio term, the boundary offset suppression term, and the time delay stability term to obtain the boundary alignment value.

7. The self-healing method for site early warning and evidence storage based on peer-to-peer edge redundancy as described in claim 1, characterized in that, The specific process of identifying the starting position and recovery interval of data recovery based on the boundary reliability assessment results, and realizing data completion and incremental recovery, is as follows: Multiply the difference between the boundary alignment value and the longest consecutive identical suffix length, and round up to get the number of boundary backs. Subtract the longest consecutive identical suffix length from the maximum warning record number of the current edge node, add one, and then subtract the number of boundary backs to get the common data boundary number. When the local maximum warning record number is less than the common data boundary number, it is determined that there is a data gap before the common data boundary in the new primary backup node; the interval between the local maximum warning record number plus one and the common data boundary number is read as the missing data filling interval, and the current edge node sends the standardized warning record, the corresponding warning record number, and the chained evidence value within the missing data filling interval to the new primary backup node; after the missing data filling is completed, the interval between the common data boundary number plus one and the current edge node's maximum warning record number is read as the incremental synchronization interval, and the standardized warning record, the corresponding warning record number, and the chained evidence value within the incremental synchronization interval are sent. When the local maximum warning record sequence number is greater than or equal to the common data boundary sequence number, it is determined that the new active-standby node has covered the data before the common data boundary, the interval between the local maximum warning record sequence number plus one and the maximum warning record sequence number of the current edge node is read as the data synchronization interval, and the current edge node sends the standardized warning records, the corresponding warning record sequence numbers and the chain certificate storage values in the data synchronization interval to the new active-standby node; After receiving the data, the new active-standby node performs storage in the order of the warning record sequence numbers, calculates the chain certificate storage values item by item based on the received standardized warning records, updates the local chain certificate storage value sequence, and returns data synchronization completion confirmation information to the current edge node.

8. The self-healing method for site early warning and evidence storage based on peer-to-peer edge redundancy as described in claim 1, characterized in that, The specific process of performing connectivity status judgment based on the heartbeat interaction result, executing data mutual backup, degradation adjustment and offline certificate storage control under different connectivity statuses is as follows: Within a fixed statistical time window, count the number of heartbeat packets sent by the current edge node to the current active-standby node and the number of successful response receptions, divide the number of successful response receptions by the number of sent heartbeat packets to obtain the connectivity C between the current edge node and the active-standby node, and compare the connectivity C with the multi-level connectivity thresholds C1 and C2; when C≥C2, the current edge node is determined to be in a normal connected state; when C1≤C<C2, the current edge node is determined to be in a connected degradation state; when C<C1, the current edge node is determined to be in an offline state; When the edge node is in a normal connected state, the generation of standardized warning records, the data mutual backup relationship between active-standby nodes and the calculation of chain certificate storage values are maintained; when the edge node is in a connected degradation state, on the premise of shortening the heartbeat detection period, the generation of standardized warning records, the data mutual backup relationship between active-standby nodes and the calculation of chain certificate storage values are maintained; when the edge node is in an offline state, sending standardized warning records, warning record sequence numbers and chain certificate storage values to the active-standby node is suspended, and only the generation of standardized warning records, the update of chain certificate storage abstract and the calculation of chain certificate storage values are continuously performed locally.

9. The self-healing method for site early warning and evidence storage based on peer-to-peer edge redundancy as described in claim 1, characterized in that, The specific process of triggering the self-healing process when the network is restored, completing the consistency recovery of chain data and the reconstruction of continuous certificate storage is as follows: When the node operating state is restored from the offline state or the connected degradation state to the normal connected state, the offline self-healing recovery process is triggered: determining the current active-standby node of the edge node, sending the maximum warning record sequence number, the corresponding chain certificate storage abstract and the unique node identifier to the active-standby node, calculating the boundary alignment value, completing the determination of the common data boundary sequence number, the judgment of the data missing interval and the data synchronization interval, and the retransmission of the corresponding standardized warning records, and updating the standardized warning record sequence and the chain certificate storage value sequence in the active-standby node.

10. A self-healing system for early warning and evidence storage at a site based on peer-to-peer edge redundancy, characterized in that: Comprising: A station perception data processing module, configured to collect multi-source operation data of edge nodes in real time, perform time sequence unification and validity screening on the multi-source operation data, form standardized warning records based on warning judgment, construct a continuous chain certificate storage sequence, and extract a chain certificate storage abstract; The peer-to-peer edge backup takeover module is used to perform communication quality assessment based on the connectivity between nodes, determine the list of primary backup nodes and candidate nodes based on the communication quality assessment, establish data backup relationship, and execute candidate takeover and takeover request sending when the primary node is unreachable. The continuous evidence storage boundary recovery module is used to perform alignment matching based on the chain evidence storage digest for the nodes under takeover, determine the consistency degree of the chain tail and the degree of boundary offset, form a boundary reliability assessment result, and identify the data recovery start position and recovery interval based on the boundary reliability assessment result to realize data completion and incremental recovery. The self-healing module for evidence storage status switching is used to determine the connectivity status based on the heartbeat interaction results, perform data backup, degradation adjustment and network outage evidence storage control under different connectivity statuses, and trigger the self-healing process when the network is restored to complete the chain-like data consistency recovery and continuous evidence storage reconstruction.

Images