Safety production monitoring data processing method and system based on edge computing

Abstract

The application discloses a kind of based on edge computing's safety production monitoring data processing method and system. Edge node receives target service area safety production monitoring data, generates link state record to determine communication degradation phase and communication recovery phase;Degradation phase is converted into event fragment carrying object anchor point, area anchor point and chain position anchor point, forms event chain skeleton, and exchanges neighborhood witness abstract with adjacent node by local edge link independent of central link;Recovery phase receives the central confirmation anchor point returned by central platform, determines recovery starting point according to event chain skeleton and neighborhood witness abstract, generates bridging fragment for the position that has not formed the relationship of succession, recovery starting point, event chain skeleton, neighborhood witness abstract and bridging fragment are packaged as difference recovery synchronization package and sent to central platform, realize the continuous connection of edge side monitoring process and central confirmed state, reduce repeated transmission and missing transmission.

Description

Technical Field

[0001] This invention relates to the field of edge computing and safety production monitoring data processing technology, specifically to a safety production monitoring data processing method and system based on edge computing, as well as a computer-readable storage medium. Background Technology

[0002] Safety monitoring systems in industrial parks, factories, and computer rooms are gradually adopting edge computing architecture. Typically, multiple edge nodes access safety monitoring data within their respective service areas, including equipment operation data, environmental status data, access control status, video summaries, and edge node processing records. Edge nodes perform local preprocessing, data aggregation, and initial status organization near the data source before synchronizing with the central platform, enabling the central platform to centrally maintain the monitoring process for each area.

[0003] When communication between edge nodes and the central platform is normal, existing safety production monitoring systems typically synchronize field data to the central platform according to the collection time, data number, or upload queue. In the event of weak network links, jitter, or short-term interruptions, the common approach is for edge nodes to cache the unuploaded data and record local logs, and then re-upload the data according to the timestamp, upload marker, or queue order after the link is restored.

[0004] However, in safety production monitoring scenarios, the same abnormal process may span multiple collection cycles, multiple monitoring objects, or multiple adjacent service areas, and the above-mentioned existing technologies have the following drawbacks.

[0005] First, the monitoring process during communication degradation lacks a structured sequence of events. Simply storing cached data or local logs makes it difficult to represent the relationship between the same monitored object before and after a state change, the relationship between state changes in the same area, and the sequential relationship between adjacent data segments. After communication is restored, the central platform can only reinterpret the process that occurred during the degradation period according to the original data order.

[0006] Secondly, the status confirmed by the central platform is difficult to align directly with the local processing of edge nodes. When retransmission is performed based solely on timestamps or upload queues after communication is restored, edge nodes cannot determine which position the central platform has confirmed. This can easily lead to repeated retransmissions from positions already confirmed by the central platform, or skipping processes that the central platform has not yet confirmed, resulting in omissions. It may also lead to retransmissions starting from inappropriate positions.

[0007] Third, there is a lack of reference for the boundary status of adjacent service areas. Safety monitoring events may cross adjacent areas, but the local cache of a single edge node can only reflect the data of its own area and cannot reflect the boundary status connection between adjacent service areas during the same event process. After communication is restored, it is difficult to prove the continuity of the cross-regional event process.

[0008] Fourth, existing breakpoint resumption methods typically rely on complete data retransmission. After communication is restored, the central platform needs to receive and process a large amount of data that is duplicated with the confirmed state. It is difficult to synchronize only the necessary state transition process, end state, and gaps during the recovery phase, resulting in low recovery efficiency and high link resource consumption.

[0009] Therefore, a secure production monitoring data processing method suitable for edge computing architecture is needed to maintain a continuous relationship between the edge-side monitoring process and the confirmed status of the central platform during communication degradation and recovery. Summary of the Invention

[0010] To address the shortcomings of existing technologies, this invention provides a safety production monitoring data processing method and system based on edge computing. Using edge nodes as the execution entity, during the communication degradation phase, continuously received safety production monitoring data is transformed into event fragments with object anchors, region anchors, and chain position anchors, forming an event chain skeleton. Neighborhood witness summaries are exchanged through local edge links independent of the central link. During the communication recovery phase, the recovery starting point is determined based on the central confirmation anchor, event chain skeleton, and neighborhood witness summaries. Bridging fragments are generated for positions without established continuity relationships, forming differential recovery synchronization packets. This ensures continuous connection between the edge-side monitoring process and the confirmed status of the central platform, reducing duplicate and missed retransmissions.

[0011] In a first aspect, the present invention provides a method for processing safety production monitoring data based on edge computing, comprising: The edge node receives safety production monitoring data within the target service area and generates a link status record between the edge node and the central platform. Based on the link status record, the communication degradation phase and the communication recovery phase are determined. During the communication degradation phase, the continuously received safety production monitoring data is converted into event fragments, and object anchors, area anchors, and chain anchors are generated for the event fragments. An event chain skeleton is formed based on the object anchor, the region anchor, and the chain position anchor, and neighborhood witness digests are exchanged with adjacent edge nodes through local edge links. The local edge links are local communication links between edge nodes, independent of the communication links between the edge nodes and the central platform. During the communication recovery phase, the central confirmation anchor point returned by the receiving central platform is used to determine the recovery starting point based on the central confirmation anchor point, the event chain skeleton, and the neighborhood witness summary. A bridging segment is generated at the location where no connection is formed between the central confirmation anchor point and the event chain skeleton. A differential recovery synchronization packet is formed by the recovery starting point, the event chain skeleton, the neighborhood witness summary, and the bridging segment. The differential recovery synchronization packet is then sent to the central platform.

[0012] In one possible implementation, during the communication degradation phase, the continuously received safety production monitoring data is converted into event fragments, and object anchors, region anchors, and chain anchors are generated for the event fragments. This includes: extracting the state changes of the monitored object, the state changes of the region to which it belongs, and the local processing state of the edge node from the safety production monitoring data; generating object anchors from the state differences of the monitored object before and after the state change; generating region anchors from the regional differences of the region to which it belongs before and after the state change; generating chain anchors for the corresponding event fragments based on the object inheritance relationship between object anchors of adjacent event fragments and the regional inheritance relationship between region anchors; and connecting the event fragments carrying the object anchors, region anchors, and chain anchors to the event chain skeleton.

[0013] In one possible implementation, the step of forming an event chain skeleton based on the object anchor, the region anchor, and the chain position anchor, and exchanging neighborhood witness summaries with adjacent edge nodes through a local edge link, includes: the edge node selecting a chain head anchor, a state transition segment, a chain tail anchor, and a broken chain position from the already connected event segments to form an event chain skeleton; identifying adjacent edge nodes that have a boundary association with the service area of ​​the edge node as witness nodes; sending a witness request carrying a boundary region identifier and a chain position anchor to the witness node; receiving a neighborhood witness summary returned by the witness node, the neighborhood witness summary carrying the boundary region status, summary generation time, corresponding chain position anchor, and witness node identifier; and attaching the neighborhood witness summary to the corresponding chain position anchor in the event chain skeleton.

[0014] In one possible implementation, the step of extracting the state changes of the monitored object, the state changes of the region to which it belongs, and the local processing status of the edge node from the safety production monitoring data includes: extracting the previous confirmed state and the current acquisition state of the monitored object from the equipment operation data, environmental status data, and edge node processing records within the same acquisition cycle; generating a state difference based on the previous confirmed state and the current acquisition state; and associating the state difference with the regional state of the region to which it belongs, the data receiving state of the edge node, and the segment processing state to obtain anchor point base data for generating object anchor points and regional anchor points.

[0015] In one possible implementation, the step of selecting the first anchor point, state transition segment, last anchor point, and broken chain position from the already connected event segments by the edge node to form an event chain skeleton includes: determining the object anchor point, region anchor point, and chain position anchor point carried by the first event segment connected to the event chain skeleton during the communication degradation phase as the first anchor point; selecting, among the already connected event segments, event segments where the state difference direction of the object anchor point changes, the region state of the region anchor point undergoes cross-regional succession, or the succession sequence of the chain position anchor point jumps as state transition segments; determining the object anchor point, region anchor point, and chain position anchor point carried by the last event segment connected to the event chain skeleton before the start of the communication recovery phase as the last anchor point; and determining the position where the chain position anchor points of adjacent event segments do not form a continuous succession sequence as a broken chain position.

[0016] In one possible implementation, receiving the neighborhood witness summary returned by the witness node includes: receiving neighborhood witness summaries returned by multiple witness nodes; filtering the neighborhood witness summaries according to the connection relationship between the boundary region state in the neighborhood witness summary and the region anchor point pointed to by the corresponding chain anchor point; retaining the neighborhood witness summaries that form a connection relationship with the region anchor point and whose summary generation time is within the collection time or collection period of the event segment to which the corresponding chain anchor point belongs; and attaching the retained neighborhood witness summaries to the corresponding chain anchor point in the event chain skeleton.

[0017] In one possible implementation, during the communication recovery phase, receiving the central confirmation anchor point returned by the central platform and determining the recovery starting point based on the central confirmation anchor point, the event chain skeleton, and the neighborhood witness summary includes: performing a connection comparison between the central confirmation anchor point and the chain head anchor point, the object anchor point and region anchor point carried by the state transition segment, and the chain tail anchor point in the event chain skeleton; verifying the connection comparison results based on the boundary region state in the neighborhood witness summary; and selecting the event segment that passes the verification and has the earliest connection sequence as the recovery starting point according to the segment connection sequence recorded in the chain anchor point.

[0018] In one possible implementation, generating bridging segments at positions where no connection is formed between the central confirmation anchor point and the event chain skeleton, and forming a differential recovery synchronization package, includes: extracting adjacent confirmable anchor points on both sides of the position where no connection is formed; generating bridging segments based on the adjacent confirmable anchor points, broken chain positions, corresponding neighborhood witness summaries, and segment sources; selecting state transition segments and chain tail segments located after the recovery start point from the event chain skeleton; and encapsulating the recovery start point, the state transition segments, the chain tail segments, the bridging segments, and the neighborhood witness summaries into a differential recovery synchronization package.

[0019] Secondly, the present invention provides a safety production monitoring data processing system based on edge computing. The system is deployed at an edge node and includes a data receiving module, a fragment generation module, a skeleton generation module, a recovery and positioning module, and a synchronization generation module. Each module is used to implement the corresponding steps of the above method.

[0020] Thirdly, the present invention provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the method described in any of the preceding claims.

[0021] Compared with existing technologies, this invention has the following advantages: First, during the communication degradation phase, continuously received data is transformed into event fragments carrying object anchors, region anchors, and chain anchors, enabling the edge-side monitoring process to have a structured continuity relationship that is recoverable, comparable, and differentially synchronized. Second, the event chain skeleton retains the chain head anchor, state transition fragments, chain tail anchors, and broken chain positions, allowing the central platform to quickly understand the key processes during degradation after recovery, without having to rely entirely on the original data queue order. Third, by exchanging neighborhood witness summaries through local edge links independent of the central link, a reference is provided for the status across service areas or boundary areas. Fourth, during the recovery phase, the central confirmation anchor, event chain skeleton, and neighborhood witness summaries are used to jointly determine the recovery starting point, reducing the possibility of repeated recovery and missed recovery. Fifth, bridging fragments are generated for positions where no continuity relationship has been formed, and differential recovery synchronization packets are formed, reducing unnecessary complete retransmissions and improving the synchronization efficiency during the recovery phase. Attached Figure Description

[0022] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the following description of the embodiments are briefly introduced. Obviously, the drawings described below are merely some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without any creative effort.

[0023] Figure 1 A flowchart illustrating the safety production monitoring data processing method based on edge computing provided in an embodiment of the present invention; Figure 2 This is a schematic diagram illustrating an application scenario of the method provided in an embodiment of the present invention, showing the communication relationship between edge nodes, adjacent edge nodes, and the central platform; Figure 3 A schematic diagram of the structure of an event fragment and its object anchor point, region anchor point and chain anchor point provided in an embodiment of the present invention; Figure 4 This is a schematic diagram of the event chain skeleton provided in an embodiment of the present invention, showing the chain head anchor point, state transition segment, chain tail anchor point, broken chain position, and the attached neighbor witness summary. Figure 5 This is a schematic diagram of the differential recovery synchronization package provided in an embodiment of the present invention; Figure 6 This is a schematic diagram of the structure of a safety production monitoring data processing system based on edge computing provided in an embodiment of the present invention. Detailed Implementation

[0024] To make the objectives, technical solutions, and advantages of the present invention clearer, the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments. It should be understood that the specific embodiments described herein are only for explaining the present invention and are not intended to limit the present invention. Where there is no conflict, the technical features of the various embodiments of the present invention can be combined with each other.

[0025] To facilitate understanding, some of the terms used in this invention will be explained first.

[0026] Furthermore, to clarify the technical boundaries of object anchors, region anchors, chain anchors, event chain skeletons, neighborhood witness summaries, bridging fragments, and differential recovery synchronization packets in this invention, the above data processing structures are all used to describe the state succession relationship of safety production monitoring data in the communication degradation and communication recovery phases, rather than being used merely as abstract naming.

[0027] Specifically, the object anchor point includes at least the monitored object identifier, previous confirmed status, current acquisition status, status difference direction, status difference generation time, and object status summary; the region anchor point includes at least the service region identifier, boundary region identifier, previous region status, current region status, region difference direction, and region status summary; the chain anchor point includes at least the preceding segment identifier, subsequent segment identifier, segment continuation sequence, continuation type, and break marker; the event chain skeleton includes at least the skeleton identifier, chain head anchor point, set of state transition segments, chain tail anchor point, and set of broken chain positions. The set of witness summaries for the neighborhood is combined; the neighborhood witness summaries include at least the witness node identifier, boundary region identifier, boundary region status, summary generation time, corresponding chain anchor point, and summary verification identifier; the bridging segment includes at least the bridging segment identifier, broken chain position, front verifiable anchor point, back verifiable anchor point, neighborhood witness summary reference, segment source, and bridging description; the differential recovery synchronization packet includes at least the synchronization packet identifier, edge node identifier, recovery start point, state transition segment set, chain tail segment, bridging segment set, neighborhood witness summary set, and original data reference set.

[0028] The type of connection in the chain anchor is determined based on the object connection relationship and the region connection relationship. When neither the object connection relationship nor the region connection relationship between adjacent event segments is valid, or the segment connection sequence is discontinuous, the break marker is set to true. The neighborhood witness summary is used to verify the connection between the boundary region state and the region anchor. The bridging segment is used to fill the connection gap based on the verifiable anchors on both sides of the broken chain position. The differential recovery synchronization packet is used to carry the necessary event segments, summaries, and reference information that need to be sent to the central platform during the recovery phase.

[0029] An edge node is a computing node deployed near the data source for safety production monitoring, responsible for data access and local processing within a target service area. The target service area refers to the physical or logical area monitored by the edge node, such as server room area 1, production line area 2, or corridor area.

[0030] The central platform refers to the computing platform that synchronizes data with each edge node and centrally maintains the monitoring process and confirmation status of each service area.

[0031] Safety production monitoring data refers to data collected or accessed by edge nodes from their target service areas. In this embodiment, it includes, but is not limited to, equipment operation data, environmental status data, access control status, video summaries, and edge node processing records. Equipment operation data includes equipment operating status, operating parameters, start / stop status, and operating condition indicators; environmental status data includes temperature and humidity, smoke, gas concentration, liquid level, pressure, and electrical status; video summaries are summaries of object locations, action states, area occupancy states, or target changes extracted by edge nodes from video streams, and are structured descriptive data rather than the original video; edge node processing records include data reception status, segment processing status, local cache status, and transmission status.

[0032] The collection period refers to the time interval during which an edge node collects data from the target service area at a fixed or configurable time interval. In this embodiment, the collection period is numbered, such as the kth collection period, or identified by a timestamp.

[0033] The following is combined Figure 1 The method for processing safety production monitoring data based on edge computing provided in this embodiment of the invention will be described. The execution subject of the method is an edge node, and it includes steps S1 to S5.

[0034] Step S1: The edge node receives safety production monitoring data within the target service area and generates a link status record between the edge node and the central platform. Based on the link status record, the communication degradation phase and the communication recovery phase are determined.

[0035] Edge nodes receive safety production monitoring data within their target service area and simultaneously generate a link status record between the edge node and the central platform. This link status record records information such as the communication status, duration of status, most recent central confirmation identifier, and most recent transmission batch identifier of the communication link between the edge node and the central platform at different times or collection cycles.

[0036] In one embodiment, the link status record includes the following fields: edge node identifier, indicating the edge node that generated the record, such as edge node A; service area identifier, indicating the target service area that the edge node is responsible for, such as data center zone 1; link status identifier, indicating the communication status, with values ​​including normal, weak network, jitter, interruption, recovery handshake; status record time, indicating the generation time or collection cycle number of the record; most recent central confirmation identifier, indicating the location of the event most recently confirmed by the central platform, which can correspond to the central confirmation anchor point described later; and local reception sequence, indicating the data sequence number continuously received locally by the edge node, used to associate with the event segment sequence number.

[0037] The communication degradation phase refers to the stage where the communication link between the edge node and the central platform is in a weak network, jittery, interrupted, or unable to continuously complete central confirmation. The communication recovery phase refers to the stage where the edge node and the central platform re-establish a communication connection and are able to receive central confirmation anchor points returned by the central platform.

[0038] One specific method for edge nodes to determine the communication degradation and recovery phases based on link status records is as follows: When the link status identifier in the link status record changes from normal to weak network, jitter, or interruption, or when the edge node sends data to the central platform but does not receive a confirmation response from the central platform within a set number of consecutive collection cycles, the edge node determines that it has entered the communication degradation phase and records the collection cycle for entering the degradation phase as the starting point of the degradation phase. The set number is a degradation determination parameter written into the link status record, which is determined by the central link timeout configuration, the on-site network level configuration, or the edge node's local configuration. If not configured, two collection cycles can be used. When the link status identifier changes from weak network, jitter, or interruption to a recovery handshake, and the edge node receives a central confirmation anchor point from the central platform again, the edge node determines that it has entered the communication recovery phase. Through the above determination, the edge node can independently divide the phase locally without relying on the central platform, providing a phase basis for subsequent event segment generation and recovery synchronization.

[0039] See Figure 2In this application scenario, edge node A is responsible for area 1 of the computer room, and edge node B is responsible for the corridor area adjacent to area 1 of the computer room. Edge node A and edge node B are connected through a local edge link, and edge nodes A and B are connected to the central platform through their respective central links. A and B are connected to the central platform through central links and communicate with each other through local edge links.

[0040] Step S2: During the communication degradation phase, the continuously received safety production monitoring data is converted into event fragments, and object anchors, area anchors, and chain anchors are generated for the event fragments.

[0041] During the communication degradation phase, the edge node continues to receive safety production monitoring data within the target service area and converts the continuously received data into event fragments. At the same time, it generates object anchors, region anchors, and chain anchors for the event fragments, as specifically implemented below.

[0042] Extracting Status Changes and Local Processing Status. Edge nodes extract the status changes of monitored objects, the status changes of their respective areas, and the local processing status of the edge nodes from the safety production monitoring data. The monitored objects refer to equipment, area targets, access control points, or video summary objects. Specifically, edge nodes extract the previous confirmed status and the current acquisition status of monitored objects from equipment operation data, environmental status data, and edge node processing records within the same acquisition cycle. The previous confirmed status refers to the most recently confirmed status value of the monitored object before the current acquisition cycle; confirmation includes confirmation by the central platform during normal communication or confirmation by a previous event segment during a degradation phase. The current acquisition status refers to the status value acquired by the monitored object within the current acquisition cycle.

[0043] Edge nodes generate state differences based on the previous confirmed state and the current acquisition state. The state difference refers to the difference result obtained by comparing the previous confirmed state and the current acquisition state of the same monitored object at the field level. It includes at least the changed state field, the value before the change, the value after the change, and the direction of the state difference. The direction of the state difference refers to the trend of the state field value change, such as "increase" or "decrease" for numerical fields, "from closed to open" or "from open to closed" for switch fields, and "from static to dynamic" or "from dynamic to static" for static / dynamic fields. For example, if the operating current field of a power distribution device changes from 12A in the previous confirmed state to 18A in the current acquisition state, then the state difference is {Field: Operating Current, Previous Value: 12A, Next Value: 18A, Direction: Increase}; similarly, if the smoke status field in area 1 of the computer room changes from "none" to "present", then the direction of the state difference is "from none to present".

[0044] Edge nodes associate the state differences with the regional state of their respective areas, the data reception state of the edge nodes, and the fragment processing state to obtain the basic anchor data used to generate object anchors and regional anchors. The data reception state indicates whether the data has been completely received within the current acquisition cycle, and the fragment processing state indicates whether the fragment has been processed and cached locally.

[0045] Generate event segments. An event segment refers to a monitoring process segment related to changes in the status of the same monitored object or service area, compiled by an edge node within one acquisition cycle or a set of consecutive acquisition cycles, based on safety production monitoring data. See also... Figure 3 In one embodiment, the safety production monitoring data is converted into event fragments. Each event fragment includes a fragment identifier, a monitored object identifier, a service area identifier, a collection time or collection cycle, a previous confirmation status, a current collection status, a local processing status, and a reference to the original data. The fragment identifier represents a unique identifier for the event fragment locally on the edge node, such as FA-001. The original data reference indicates the local storage location of the corresponding original data or summary data, for subsequent retrieval and verification by the central platform. The event fragment is further associated with object anchors, area anchors, and chain position anchors, which respectively represent differences in the monitored object status, differences in the service area status, and the sequence position of the event fragment in the event chain skeleton.

[0046] Object anchor points are generated. Edge nodes generate object anchor points based on the state differences of monitored objects before and after state changes. An object anchor point is a structured identifier representing the state differences of a monitored object before and after a state change. In one embodiment, it consists of the monitored object identifier, the previous confirmed state, the current acquisition state, the direction of the state difference, the state difference generation time, and the object state summary. The object state summary is a summary value calculated after serializing the above fields, used for rapid comparison and integrity verification.

[0047] Generate region anchor points. Edge nodes generate region anchor points by representing the regional differences of their respective regions before and after a state change. A region anchor point is a structured identifier representing the regional differences of its own region or boundary region before and after a state change. In one embodiment, it consists of a service region identifier, a boundary region identifier, the previous region state, the current region state, the direction of the regional difference, and a region state summary. The boundary region identifier is used to identify adjacent regions that have a boundary association with this service region.

[0048] Generate chain anchor points. Edge nodes generate chain anchor points for corresponding event segments based on the object succession relationships between object anchor points of adjacent event segments and the region succession relationships between region anchor points. A chain anchor point is a structured identifier representing the succession order of an event segment in the event chain skeleton and its preceding and following succession relationships. In one embodiment, it includes a preceding segment identifier, a subsequent segment identifier, a segment succession order, a succession type, and a break marker. The segment succession order refers to a monotonically increasing sequence number assigned by the edge node to each event segment accessing the event chain skeleton during the communication degradation phase, such as 1, 2, 3…

[0049] The method for determining the object inheritance relationship is as follows: For adjacent preceding and following event segments, if their monitored object identifiers are the same, and the previous confirmation state of the object anchor point of the following event segment is consistent with the current collection state of the object anchor point of the preceding event segment, then the object inheritance relationship between the two is determined to be established. The method for determining the region inheritance relationship is as follows: For adjacent preceding and following event segments, if their service region identifiers are the same or there is a boundary association, and the status field dimension, collection period or time window, and data type corresponding to their region anchor points are consistent, and the previous region state of the region anchor point of the following event segment is consistent with the current region state of the region anchor point of the preceding event segment, then the region inheritance relationship between the two is determined to be established. The inheritance type is determined based on the above determination results as object inheritance, region inheritance, or both object and region inheritance. When neither object inheritance nor region inheritance is established between adjacent event segments, or the segment inheritance sequence is not continuous, the break flag of the chain anchor point is set to true.

[0050] Edge nodes will connect event fragments carrying object anchors, region anchors, and chain position anchors to the event chain skeleton, thereby transforming the continuous monitoring data generated during communication degradation into a recoverable, comparable, and differentially synchronized event stream structure.

[0051] Step S3: Based on the object anchor point, the region anchor point and the chain position anchor point, form an event chain skeleton, and exchange neighborhood witness digests with adjacent edge nodes through the local edge link. The local edge link is a local communication link between edge nodes, independent of the communication link between the edge node and the central platform.

[0052] Step S3 specifically includes: An event chain skeleton is formed. The event chain skeleton refers to the event flow structure consisting of the chain head anchor, state transition segments, chain tail anchor, broken chain positions, and neighborhood witness summaries. The event chain skeleton is not a complete copy of all the original data during communication degradation, but rather retains key segments, anchors, and broken positions used for synchronization recovery. See also... Figure 4Edge nodes select the chain start anchor, state transition segment, chain end anchor, and broken chain position from the already connected event segments, as follows: The object anchor point, region anchor point, and chain position anchor point carried by the event fragment of the first access event chain skeleton during the communication degradation phase are determined as the chain head anchor point. Among the connected event segments, select the event segment that meets any of the following conditions as the state transition segment: The state difference direction of the object anchor point changes relative to the previous event segment, for example, from "increase" to "decrease", or from "from static to dynamic" to "from dynamic to static"; The region anchor point and the neighbor witness summary of the adjacent service region form a cross-regional connection under the same state field dimension; The connection sequence of the chain anchor point jumps relative to the previous event segment, that is, the connection sequence is not continuously increasing. The object anchor, region anchor, and chain position anchor carried by the event fragment of the last access event chain skeleton before the start of the communication recovery phase are determined as the chain tail anchor. The position where a continuous successor sequence is not formed between the chain anchor points of adjacent event segments is determined as a broken chain position, that is, the position where the subsequent segment identifier of the previous event segment does not correspond to the preceding segment identifier of the next event segment, or the difference in their successor sequence is greater than 1.

[0053] In one embodiment, the event chain skeleton includes the following fields: skeleton identifier, chain head anchor, set of state transition segments, chain tail anchor, set of broken chain positions, and set of neighborhood witness summaries.

[0054] Witness nodes are identified. Edge nodes identify adjacent edge nodes that have a boundary association with their service area as witness nodes. This boundary association is derived from service area configuration, physical adjacency, monitored object flow relationships, or area linkage configuration. For example, edge node A, responsible for area 1 of the computer room, identifies edge node B, responsible for the adjacent corridor area, as a witness node.

[0055] Exchange neighboring witness summaries. Edge nodes send witness requests carrying boundary area identifiers and chain anchors to witness nodes via local edge links. The local edge link refers to a local communication link between edge nodes, independent of the communication link between the edge node and the central platform. Therefore, even if communication degradation occurs between the edge node and the central platform, adjacent edge nodes can still exchange summary information within their local network. In one embodiment, the local edge link is implemented using a local area network, industrial Ethernet, field wireless network, or a dedicated communication channel between edge nodes.

[0056] The witness node generates and returns a neighborhood witness summary based on the witness request. The neighborhood witness summary refers to the summary information generated by adjacent edge nodes for the boundary region status, summary generation time, and corresponding chain anchor point. It carries the boundary region status, summary generation time, corresponding chain anchor point, and witness node identifier. In one embodiment, the neighborhood witness summary includes the following fields: witness node identifier, boundary region identifier, boundary region status, summary generation time, corresponding chain anchor point, and summary verification identifier. The summary verification identifier is used for summary integrity verification or duplicate summary identification.

[0057] Filter and attach neighbor witness summaries. Edge nodes may receive neighbor witness summaries returned by multiple witness nodes. First, perform integrity checks and duplicate summary identification on the neighbor witness summaries based on the summary verification identifier, and remove neighbor witness summaries that fail the verification or are duplicates. Then, filter them as follows: match the boundary region identifier, state field dimension, summary generation time and data type in the neighbor witness summary with the region anchor point pointed to by the corresponding chain anchor point; when the boundary region identifiers correspond, the state field dimensions are consistent, the summary generation time is within the collection time or collection period of the event segment to which the corresponding chain anchor point belongs, and the boundary region state and the current region state of the region anchor point are connected at the boundary, it is determined that the two form a connection; among them, numerical state fields are compared when the units are consistent or after conversion by a preset unit, switch or enumeration state fields are compared when the value sets are consistent, and video summaries are only compared with their structured summary fields rather than the original video content; retain the neighbor witness summaries that meet the above matching and connection conditions. Edge nodes will attach the preserved neighborhood witness summaries to the corresponding chain anchor points in the event chain skeleton, thereby providing boundary state references from adjacent edge nodes for cross-regional event processes on the event chain skeleton. Neighborhood witness summaries that do not meet the above conditions, such as those whose summary generation time falls outside the corresponding acquisition period, or whose boundary region state cannot be connected with the region anchor point, will be discarded to avoid introducing irrelevant or outdated boundary references.

[0058] Step S4: During the communication recovery phase, the central confirmation anchor point returned by the central platform is received, and the recovery starting point is determined based on the central confirmation anchor point, the event chain skeleton, and the neighborhood witness summary.

[0059] During the communication recovery phase, after the edge node re-establishes communication with the central platform, it receives a central confirmation anchor point returned by the central platform. This central confirmation anchor point is a structured identifier returned by the central platform that indicates the location of the monitored state it has confirmed. It represents the monitoring state location that the central platform confirmed before communication degradation or during the recovery handshake. In one embodiment, the central confirmation anchor point includes the following fields: central confirmation identifier, confirmed object status, confirmed area status, confirmation time or confirmation period, and confirmation batch identifier.

[0060] Edge nodes determine the recovery starting point based on the central confirmation anchor, event chain skeleton, and neighborhood witness summary, specifically including: The process involves several steps. First, edge nodes determine candidate event segments from the event chain skeleton based on the confirmed object status, confirmed area status, confirmation time or period, and confirmation batch identifier carried by the central confirmation anchor. Then, they perform a connection comparison between the central confirmation anchor and the corresponding chain head anchor, the object anchor and area anchor carried by the state transition segment, and the chain tail anchor for each candidate event segment. This connection comparison includes object-side comparison, area-side comparison, and chain position-side comparison: Object-side comparison determines whether the confirmed object status of the central confirmation anchor can be connected to the previous confirmed status or the current acquisition status of the event segment object anchor. That is, if the confirmed object status is consistent with the previous confirmed status under the premise of the same monitored object identifier, the object side is considered to be able to connect. Area-side comparison determines whether the confirmed area status of the central confirmation anchor can form a continuous relationship with the area status change represented by the area anchor within the same status field dimension and acquisition time window. Chain position-side comparison determines whether the segment connection sequence can continue continuously from the central confirmation position, excluding event segments whose confirmation time or period is earlier than or equal to the central confirmed position and which have already been confirmed.

[0061] Verification. Edge nodes verify the acceptance comparison results based on the boundary region status in the neighborhood witness summary. For event procedures involving adjacent service regions, edge nodes use the neighborhood witness summary attached to the corresponding chain anchor point to confirm whether the boundary region status is connected to the region anchor point pointed to by the corresponding chain anchor point; if they are connected, the acceptance comparison result passes the verification; otherwise, it fails the verification.

[0062] A recovery starting point is selected. Edge nodes arrange event segments that have passed verification and are not covered by the centrally confirmed anchor point according to the segment succession sequence recorded in the chain anchor points, and select the event segment with the earliest succession sequence as the recovery starting point. The recovery starting point refers to the event segment or anchor point position determined during the communication recovery phase to begin differential recovery synchronization. Therefore, the recovery starting point is not simply the first cached data after the communication interruption, but is determined jointly based on the centrally confirmed anchor point, the local event chain skeleton, and the neighborhood witness digest. Starting before a centrally confirmed position will result in duplicate recovery; skipping processes not yet confirmed by the central will result in missed recovery. This method reduces the risk of duplicate and missed recovery.

[0063] Step S5: Generate bridging segments for positions where no connection is formed between the central confirmation anchor point and the event chain skeleton, and form a differential recovery synchronization packet from the recovery start point, the event chain skeleton, the neighborhood witness summary, and the bridging segments, and send the differential recovery synchronization packet to the central platform.

[0064] See Figure 5 After determining the recovery starting point, the edge nodes further identify the positions where no connection relationship is formed between the central confirmation anchor point and the event chain skeleton. That is, the positions where the central confirmation anchor point cannot be continuously connected with any anchor point in the event chain skeleton in the connection comparison in step S4, or where there is a broken chain position inside the event chain skeleton, and are processed in the following manner.

[0065] Extract adjacent verifiable anchor points. For edge nodes, extract adjacent verifiable anchor points on both sides of the position where no connection relationship has been formed. These adjacent verifiable anchor points refer to anchor points located before and after the break point whose state has been confirmed or can be referenced for confirmation. These include center confirmation anchor points, chain head anchor points, object anchor points and region anchor points carried by state transition segments, chain tail anchor points, or boundary state anchor points that can be used as references in the neighborhood witness summary.

[0066] Generate bridging fragments. Edge nodes generate bridging fragments based on the adjacent verifiable anchor points, break chain positions, corresponding neighborhood witness summaries, and fragment sources. Specifically, this includes: identifying the preceding and subsequent fragment identifiers corresponding to the break chain position; extracting the current state of the anchor point before the break position and the previous state of the anchor point after the break position; determining the gap type according to object and region continuity relationships; when there are neighborhood witness summaries with the same state field dimension and matching time windows, writing the neighborhood witness summary as the bridging basis into the neighborhood witness summary reference; when no matching summary exists, marking the bridging basis as the local break cause and fragment source; and writing a bridging description based on the gap type, verifiable anchor points on both sides, neighborhood witness summary references, and fragment source. The bridging fragment refers to a supplementary fragment used to describe the position where there is a lack of direct continuity between the central verifiable anchor point and the event chain skeleton. In one embodiment, a bridging segment includes the following fields: bridging segment identifier, break link position, preceding verifiable anchor point, following verifiable anchor point, neighbor witness summary reference, and segment source, indicating that the bridging segment originates from a local event segment, a neighbor summary, or a centrally confirmed status; and a bridging description, used to record the type of connection relationship supplemented by the bridging segment. By providing verifiable statuses on both sides of the break point and a neighbor witness summary to support the judgment, the bridging segment enables the central platform to establish a referential connection at locations lacking a direct connection relationship.

[0067] Encapsulating a differential recovery synchronization packet. Edge nodes select state transition segments and chain tail segments from the event chain skeleton located after the recovery start point, and encapsulate the recovery start point, state transition segments, chain tail segments, bridging segments, and neighborhood witness summaries into a differential recovery synchronization packet. This differential recovery synchronization packet refers to a recovery synchronization data packet encapsulated from the recovery start point, necessary segments in the event chain skeleton, neighborhood witness summaries, and bridging segments. It is used to synchronize the necessary event process structure during communication degradation to the central platform, rather than retransmitting all cached data. In one embodiment, the differential recovery synchronization packet includes the following fields: synchronization packet identifier, edge node identifier, recovery start point, set of state transition segments, chain tail segment, set of bridging segments, set of neighborhood witness summaries, and set of original data references. The original data reference set is used by the central platform to retrieve or verify the original data as needed.

[0068] The edge node sends the differential recovery synchronization packet to the central platform. Upon receiving it, the central platform continues the event process formed by the edge node during the communication degradation period to the centrally confirmed state based on the recovery starting point. It also processes the gap between the centrally confirmed anchor point and the event chain skeleton according to the bridging fragment and the neighborhood witness summary, thereby completing the continuous connection between the edge-side monitoring process and the central platform's confirmed state.

[0069] The following uses a factory safety production monitoring system as an example to fully explain the above method. Multiple edge nodes are set up within the factory area. Edge node A is responsible for the computer room area 1, accessing the monitoring data of the power distribution equipment identified as D-01, including operating data, temperature and humidity data, smoke status, access control status, and video summaries. Edge node B is responsible for the corridor area adjacent to computer room area 1, communicating with edge node A via a local edge link. The central platform is responsible for the centralized maintenance of the entire factory area monitoring process. The data collection cycle is once every 10 seconds, identified by the number k.

[0070] During normal communication, edge node A continuously sends processed monitoring data to the central platform, and the central platform returns its confirmed central confirmation status. By the 100th acquisition cycle k=100, the central platform confirms that the operating current of power distribution equipment D-01 is 12A and the smoke status in area 1 of the computer room is zero. Edge node A records the most recent central confirmation identifier in the link status record.

[0071] Starting from k=101, a weak network appears between edge node A and the central platform. The degradation judgment parameter in the link status record is two collection cycles. After edge node A sends data for two consecutive collection cycles without receiving an acknowledgment response, edge node A determines that it has entered the communication degradation phase and sets k=101 as the starting point of the degradation phase. At this time, edge node A continues to receive data from the first area of ​​the equipment room and generate event fragments, as detailed below.

[0072] At k=101, the operating current of power distribution equipment D-01 increases from 12A to 14A. Edge node A generates an object anchor point, with the object being D-01, the previous value being 12A, the subsequent value being 14A, and the direction being upward. The area anchor point records that the smoke status of Zone 1 of the equipment room is still zero. This segment is the first access segment in the degradation phase, and its anchor point combination is determined as the chain head anchor point, with the segment acceptance sequence being 1.

[0073] Between k=102 and k=104, the operating current continues to rise to 18A. The connection between each segment and the previous segment is established, that is, the previous confirmation state of the next segment is equal to the current acquisition state of the previous segment, and the connection sequence is 2, 3, and 4 consecutively.

[0074] At k=105, the smoke state changes from nothing to something, the regional difference direction of the regional anchor point changes from nothing to something, and the direction of the object anchor point remains upward; since the regional state of this segment changes significantly relative to the previous sequence, it is selected as the state transition segment, with a succession position of 5.

[0075] At k=106, due to a brief packet loss during local acquisition, edge node A did not form a valid event fragment, making it impossible to directly establish object and region connections between k=105 and k=107. The break mark of the chain anchor point was set to true, forming a broken chain position.

[0076] Between k=107 and k=109, the operating current changes from increasing to decreasing, dropping from 18A to 15A; among them, the state difference direction of the object anchor point in the k=107 segment changes relative to the previous sequence, and is selected as the state transition segment.

[0077] Edge node A forms an event chain skeleton from the above fragments, retaining the chain head anchor point k=101, the state transition fragments k=105 and k=107, the broken chain position located between k=105 and k=107, and the subsequent chain tail anchor point.

[0078] During the degradation phase, edge node A sends the boundary region identifier and corresponding link anchor point between server room area 1 and the corridor area to edge node B via the local edge link. Edge node B generates a neighborhood witness summary for the corridor area. One summary's boundary region status reflects the change in smoke status from none to present or an increase in smoke concentration in the corridor area at time k=105. The summary generation time falls within the k=105 collection period. Edge node A retains this summary after filtering. Its boundary region identifier corresponds to the corridor area, the summary generation time is within the corresponding collection period, and the smoke status field and the area anchor point are mutually connected under the same environmental state dimension. It is then attached to the link anchor point corresponding to the k=105 segment. The other summary, whose generation time falls outside k=110, is discarded.

[0079] At k=110, edge node A resumes communication with the central platform, receives the recovery handshake, and receives the central confirmation anchor point. The confirmed object status of this anchor point is D-01 operating current 12A, and the confirmed area status is no smoke in server room zone 1, corresponding to the status at k=100. Edge node A compares the central confirmation anchor point with the chain head anchor point at k=101: the monitored object identifiers are the same, and the previous confirmation status 12A of segment k=101 is consistent with the confirmed object status 12A of the central confirmation anchor point. The object side can be accepted, and the area side confirmation status is also consistent. Therefore, segment k=101 passes the acceptance comparison. Edge node A then uses the neighborhood witness summary to verify that there are no conflicts in the boundary status of adjacent corridor areas in the smoke status field, collection period, and boundary area identifier. Segment k=101 passes the verification. Since k=101 has the earliest acceptance sequence, edge node A determines segment k=101 as the recovery starting point.

[0080] For the broken chain position between k=105 and k=107, edge node A extracts the adjacent verifiable anchor points on both sides, namely the object anchor point and region anchor point of the state transition segment at k=105, and the object anchor point of the state transition segment at k=107. Edge node A identifies the preceding segment of this broken chain position as k=105 and the subsequent segment as k=107, and uses the neighborhood witness summary attached to the anchor point of the chain position at k=105 as the basis for region-side bridging, and uses the local packet loss record as the fragment source to generate a bridging fragment. The bridging description record of this bridging fragment is recorded here to fill the gap between object and region acceptance caused by local packet loss.

[0081] Subsequently, edge node A selects state transition segments k=105, k=107, and chain tail segment k=109 from the event chain skeleton, located after the recovery start point k=101. It encapsulates the recovery start point, the aforementioned state transition segments, chain tail segment, bridging segment, and neighborhood witness summary into a differential recovery synchronization packet, and writes the original data reference along with it, sending it to the central platform. Based on this, the central platform continues the event process from the confirmed state at k=100, receiving only the necessary state transition segments, chain tail segment, and bridging segment, without needing to repeatedly receive all the original data from k=101 to k=109.

[0082] See Figure 6 This invention also provides a safety production monitoring data processing system based on edge computing. The system is deployed at an edge node and includes a data receiving module M1, a fragment generation module M2, a skeleton generation module M3, a recovery and positioning module M4, and a synchronization generation module M5.

[0083] The data receiving module M1 is used to receive safety production monitoring data within the target service area from the edge node, generate a link status record between the edge node and the central platform, and determine the communication degradation stage and the communication recovery stage based on the link status record. The implementation method is described in step S1.

[0084] The fragment generation module M2 is used to convert the continuously received safety production monitoring data into event fragments during the communication degradation phase, and to generate object anchors, area anchors and chain anchors for the event fragments. The implementation method is described in step S2.

[0085] The skeleton generation module M3 is used to form an event chain skeleton based on the object anchor point, the region anchor point and the chain position anchor point, and exchange neighborhood witness digests with adjacent edge nodes through the local edge link. The local edge link is a local communication link between edge nodes, which is independent of the communication link between the edge node and the central platform. Its implementation method is described in step S3.

[0086] The recovery positioning module M4 is used to receive the central confirmation anchor point returned by the central platform during the communication recovery phase, and determine the recovery starting point based on the central confirmation anchor point, the event chain skeleton and the neighborhood witness summary. The implementation method is described in step S4.

[0087] The synchronization generation module M5 is used to generate bridging segments at positions where no connection relationship is formed between the central confirmation anchor point and the event chain skeleton, and to form a differential recovery synchronization package from the recovery starting point, the event chain skeleton, the neighborhood witness summary and the bridging segments, and to send the differential recovery synchronization package to the central platform. The implementation method is described in step S5.

[0088] In the above system embodiments, the information interaction and execution process between the modules are based on the same concept as in the method embodiments of the present invention. Their specific functions and technical effects can be found in the method embodiments section, and will not be repeated here.

[0089] Storage Media Examples This invention also provides a computer-readable storage medium storing a computer program. When executed by a processor, the computer program implements the edge computing-based security production monitoring data processing method described in any of the above embodiments. The computer-readable storage medium includes, but is not limited to, 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.

[0090] It should be noted that the above embodiments are merely examples, and there are many alternative implementation methods for this invention.

[0091] First, the types of safety production monitoring data can be adjusted according to the on-site deployment. In addition to equipment operation data, environmental status data, access control status and video summaries, it can also include inspection records, work order status, energy consumption status, fire protection and weak current equipment status or edge node resource status.

[0092] Second, object anchors, region anchors, and chain anchors can be implemented using hash digests, combinations of structured fields, serialization identifiers, or database record identifiers, as long as they can express state differences and sequential positions.

[0093] Third, the local edge link can be implemented using a local area network, industrial Ethernet, field wireless network, or a dedicated communication channel between edge nodes, but it should be independent of the communication link between the edge node and the central platform.

[0094] Fourth, the neighborhood witness summary can be generated by a single adjacent edge node, or it can be generated by multiple adjacent edge nodes and then filtered and retained by the current edge node.

[0095] Fifth, the differential recovery synchronization package can carry only the necessary fragments and summaries from the event chain skeleton, or it can carry original data references for verification by the central platform. The central platform can further request supplementary data based on the original data references.

[0096] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for processing safety production monitoring data based on edge computing, characterized in that, include: The edge node receives safety production monitoring data within the target service area and generates a link status record between the edge node and the central platform. Based on the link status record, the communication degradation phase and the communication recovery phase are determined. During the communication degradation phase, the continuously received safety production monitoring data is converted into event fragments, and object anchors, area anchors, and chain anchors are generated for the event fragments. An event chain skeleton is formed based on the object anchor, the region anchor, and the chain position anchor, and neighborhood witness digests are exchanged with adjacent edge nodes through local edge links. The local edge links are local communication links between edge nodes, independent of the communication links between the edge nodes and the central platform. During the communication recovery phase, the central confirmation anchor point returned by the receiving central platform is used to determine the recovery starting point based on the central confirmation anchor point, the event chain skeleton, and the neighborhood witness summary. A bridging segment is generated at the location where no connection is formed between the central confirmation anchor point and the event chain skeleton. A differential recovery synchronization packet is formed by the recovery starting point, the event chain skeleton, the neighborhood witness summary, and the bridging segment. The differential recovery synchronization packet is then sent to the central platform.

2. The safety production monitoring data processing method based on edge computing according to claim 1, characterized in that, During the communication degradation phase, the continuously received safety production monitoring data is converted into event fragments, and object anchors, region anchors, and chain anchors are generated for the event fragments, including: Extract the status changes of the monitored objects, the status changes of the areas to which they belong, and the local processing status of the edge nodes from the safety production monitoring data; Generate object anchor points by showing the state differences of the monitored object before and after the state change; Generate regional anchor points based on the regional differences before and after the state change of the region; Based on the object inheritance relationship between object anchors of adjacent event segments and the region inheritance relationship between region anchors, chain anchors are generated for the corresponding event segments. The event fragment carrying the object anchor point, the region anchor point, and the chain position anchor point is connected to the event chain skeleton.

3. The safety production monitoring data processing method based on edge computing according to claim 1, characterized in that, The process of forming an event chain skeleton based on the object anchor, the region anchor, and the chain position anchor, and exchanging neighborhood witness summaries with adjacent edge nodes through local edge links, includes: The edge nodes select the first anchor point, state transition segment, last anchor point, and broken chain position from the already connected event segments to form the event chain skeleton; Adjacent edge nodes that have a boundary association with the service area of ​​the edge node are identified as witness nodes; Send a witness request carrying the boundary region identifier and chain anchor point to the witness node; Receive the neighborhood witness summary returned by the witness node. The neighborhood witness summary carries the boundary region status, summary generation time, corresponding chain anchor point and witness node identifier. The neighborhood witness summary is attached to the corresponding chain anchor point in the event chain skeleton.

4. The safety production monitoring data processing method based on edge computing according to claim 2, characterized in that, The extraction of the status changes of the monitored objects, the status changes of their respective regions, and the local processing status of edge nodes from the safety production monitoring data includes: Extract the previous confirmed status and current acquisition status of the monitored object from the equipment operation data, environmental status data and edge node processing records within the same acquisition cycle; A state difference is generated based on the previous confirmation state and the current acquisition state; The state differences are associated with the regional state of the region, the data receiving state of the edge nodes, and the fragment processing state to obtain the anchor point base data used to generate object anchor points and regional anchor points.

5. The safety production monitoring data processing method based on edge computing according to claim 3, characterized in that, The process of selecting the chain start anchor point, state transition segment, chain end anchor point, and broken chain position from the connected event segments by the edge node to form the event chain skeleton includes: The object anchor point, region anchor point, and chain position anchor point carried by the event fragment of the first access event chain skeleton during the communication degradation phase are determined as the chain head anchor point. Among the already connected event segments, select the event segments where the state difference direction of the object anchor point changes, the regional state of the regional anchor point is taken over across regions, or the sequence of the chain anchor point jumps as state transition segments. The object anchor, region anchor, and chain position anchor carried by the event fragment of the last access event chain skeleton before the start of the communication recovery phase are determined as the chain tail anchor. The position where a continuous sequence of anchor points is not formed between adjacent event segments is identified as a broken chain position.

6. The safety production monitoring data processing method based on edge computing according to claim 3, characterized in that, The process of receiving the neighborhood witness digest returned by the witness node, wherein the neighborhood witness digest carries the boundary region status, digest generation time, corresponding chain anchor point, and witness node identifier, including: Receive neighborhood witness summaries returned by multiple witness nodes; The neighborhood witness summary is filtered according to the connection relationship between the boundary region status in the neighborhood witness summary and the region anchor point pointed to by the corresponding chain anchor point; Retain neighboring witness summaries that form a connection with the regional anchor point and whose summary generation time is within the collection time or collection period of the event segment to which the corresponding chain anchor point belongs; The retained neighborhood witness summary is attached to the corresponding chain anchor point in the event chain skeleton.

7. The safety production monitoring data processing method based on edge computing according to claim 3, characterized in that, In the communication recovery phase, the central confirmation anchor point returned by the central platform is received, and the recovery starting point is determined based on the central confirmation anchor point, the event chain skeleton, and the neighborhood witness summary, including: The central confirmation anchor point is compared with the chain head anchor point, the object anchor point and region anchor point carried by the state transition segment, and the chain tail anchor point in the event chain skeleton. The acceptance ratio is verified based on the boundary region status in the neighborhood witness summary. Based on the sequence of events recorded in the chain anchor point, the event segment that passes the verification and has the earliest sequence is selected as the recovery starting point.

8. The safety production monitoring data processing method based on edge computing according to claim 7, characterized in that, The step of generating a bridging segment at the location where no connection is formed between the central confirmation anchor point and the event chain skeleton, and forming a differential recovery synchronization package from the recovery start point, the event chain skeleton, the neighborhood witness summary, and the bridging segment, includes: Extract the adjacent verifiable anchor points on both sides of the location where no connection relationship has been formed; A bridging fragment is generated based on the adjacent verifiable anchor points, broken chain positions, corresponding neighborhood witness summaries, and fragment sources. Select the state transition segment and the chain tail segment located after the recovery starting point from the event chain skeleton; The recovery start point, the state transition segment, the chain tail segment, the bridging segment, and the neighborhood witness summary are encapsulated into a differential recovery synchronization package.

9. A safety production monitoring data processing system based on edge computing, characterized in that, The system is deployed at edge nodes and includes: The data receiving module is used to receive safety production monitoring data within the target service area from the edge node, generate a link status record between the edge node and the central platform, and determine the communication degradation stage and the communication recovery stage based on the link status record. The fragment generation module is used to convert continuously received safety production monitoring data into event fragments during the communication degradation phase, and to generate object anchors, area anchors and chain anchors for the event fragments. The skeleton generation module is used to form an event chain skeleton based on the object anchor point, the region anchor point and the chain position anchor point, and exchange neighborhood witness digests with adjacent edge nodes through local edge links. The local edge links are local communication links between edge nodes, independent of the communication links between the edge nodes and the central platform. The recovery positioning module is used to receive the central confirmation anchor point returned by the central platform during the communication recovery phase, and determine the recovery starting point based on the central confirmation anchor point, the event chain skeleton and the neighborhood witness summary. The synchronization generation module is used to generate bridging segments at positions where no connection is formed between the central confirmation anchor point and the event chain skeleton, and to form a differential recovery synchronization package from the recovery starting point, the event chain skeleton, the neighborhood witness summary and the bridging segments, and to send the differential recovery synchronization package to the central platform.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program, which, when executed by a processor, implements the edge computing-based safety production monitoring data processing method according to any one of claims 1 to 8.

Images