CN-121301438-B - Industrial data link access method and system

Abstract

The invention discloses an on-chain access method and system for industrial data, and relates to the field of industrial data processing, wherein the method comprises the steps of collecting time sequence monitoring data of industrial equipment, arranging a trend detector at the edge side, and dynamically adjusting the anchoring frequency on a chain according to the mutation strength and abnormal confidence coefficient of the data; and when the transient event is captured, generating an event snapshot containing a time window and an associated point location, preferentially uploading the event snapshot through an independent channel, and falling back to the conventional anchoring rhythm after abnormal fading. When the historical data is archived in a split mode, a linking record containing an inter-volume association index is generated in a linking mode through volume switching, a cross-volume index catalog corresponding to the link Ha Xigen is maintained, and an anchoring time period window on the link is inquired before the archiving system cuts the volume to align the boundary. The method solves the problems of transient abnormal tracing blind area and cross-roll chain breakage caused by the existing fixed anchoring, improves the data tracing consistency and the key event capturing capability, and provides reliable support for industrial data audit and fault tracing.

Inventors

• ZHOU XIAOHUI
• ZOU XIAOYONG
• ZHU XINHE

Assignees

• 盐城数融智升科技有限公司

Dates

Publication Date

20260512

Application Date

20251022

Claims (6)

1. 1. An on-industrial data chain access method, comprising the steps of: S1, acquiring time-amplitude original data from industrial data chain monitoring points, and constructing a two-dimensional coordinate system according to the time-amplitude original data, wherein unit judgment time slot parameters are set in the two-dimensional coordinate system, time slot sliding algorithm is input into the time slot parameters, time slot window data are output, and the time slot window data are subjected to non-maximum suppression to obtain a suppressed anchor sequence; S2, inputting the inhibited anchor sequence into a relevant fragment type judging algorithm to obtain high-relevance anchor point data, and obtaining bookmark total table data after time interval expansion, hash value extraction, time correction and aggregation index of the high-relevance anchor point data; S3, performing end-to-end butt joint verification by applying bookmark total table data to generate inter-volume joint record data, and performing time sequence arrangement after performing missing segment supplementary recording on the inter-volume joint record data to generate complete cross-volume directory data; S4, constructing two visual views according to the complete cross-volume directory data, generating a continuity certificate based on the two views, inputting a time interval threshold value to detect and obtain a gap mark, and rechecking the gap mark to obtain a rechecking result; The step of inputting the high-association anchor point data by the association fragment type judgment algorithm into the anchor sequence after suppression comprises the following steps: Constructing an associated fragment type judging algorithm, inputting an anchor sequence after inhibition, defining a maximum candidate time difference, a similarity weight, an associated threshold and a type judging threshold group, wherein the associated fragment type judging algorithm firstly calculates a lightweight fingerprint in a window feature vector of the anchor sequence after inhibition, screens candidate anchor pairs by using a time proximity constraint and a scoring significance constraint, and outputs candidate pairs; Calculating five-dimensional similarity for each candidate pair, weighting and synthesizing association scores, constructing an undirected graph by taking the suppressed anchor sequence as a node and taking the anchor pair with the association score as an edge, and carrying out connected domain on the undirected graph or carrying out community detection based on modularity to obtain an anchor cluster; Calculating anchor point cluster level characteristics for each anchor point cluster, judging the type of the associated fragment, and outputting the type and the confidence; the obtaining bookmark total data by expanding the time interval, extracting the hash value, correcting the time and aggregating the index of the high-association anchor point data comprises the following steps: the high-association anchor point data is based on an anchor point cluster set output by the anchor point clusters, and self-adaptive expansion is carried out on the original time range of each anchor point cluster; pairing each anchor point cluster with the corresponding expansion fragment to obtain an expansion fragment set, and performing strong hash extraction on the expansion fragment set to obtain a hash chain set corresponding to each anchor point cluster; A time correction model is led out from the expansion fragment set and the anchor sequence after inhibition and is applied, an affine drift model is adopted to describe the mapping relation between equipment time and reference time, and each sample time stamp in the expansion fragment is corrected according to the affine drift model, so that a block-level tuple set with anchor point cluster numbers and block numbers as indexes is obtained; Constructing a multi-type index by using a block-level tuple set, wherein the multi-type index comprises a time index, a hash index and an anchor context index, obtaining an aggregation index after the multi-type index, organizing each block into uniform bookmark entries according to the aggregation index, and forming bookmark total data according to all the bookmark entries; Performing end-to-end butt joint verification on the bookmark summary data to generate inter-volume linking record data, performing time sequence arrangement on the inter-volume linking record data after performing missing segment complement, wherein generating complete cross-volume directory data comprises: The method comprises the steps of performing butt joint judgment on two adjacent volumes of a bookmark item by using bookmark total table data, and calculating a candidate pair consistency score; Searching anchor point clusters with overlapping time ranges with gap intervals after expansion in high-association anchor point data, and taking the anchor point clusters with types being re-recorded or sequentially extended as priority candidate anchor point clusters; Inserting the supplementary bookmark items into the bookmark total data, and then ordering again according to the starting time after timing, after ordering, distributing a global serial number for each bookmark item, wherein the serial number of each global serial number is equal to the position serial number of the bookmark item corresponding to the global serial number in the bookmark total data; constructing two visual views according to the complete cross-volume directory data, generating a continuity certificate based on the two views, inputting a time interval threshold value to detect and obtain a gap mark, and rechecking the gap mark to obtain a rechecking result, wherein the method comprises the following steps of: leading out a time axis view and a topology view from the complete cross-volume directory data, wherein the time axis view and the topology view are commonly used for generating a continuity certificate and a rechecking record; constructing a leaf hash for each bookmark item, merging the leaf hash layer by layer to obtain a hash root, and digitally signing Ha Xigen and the time range covered by the complete cross-volume directory by using a private key to form a continuity certificate; Calculating adjacent intervals of two adjacent bookmark items in the bookmark total table data, calculating a time interval threshold according to the adjacent intervals, and marking positions between the corresponding adjacent bookmark items as gaps when a certain adjacent interval is larger than the time interval threshold; and rechecking the position marked as the notch according to three situations, including the situation that the evidence is sufficient but a service empty window exists, the situation that the repairing of the bookmark entry can be carried out by the supplementary record and the situation that the butt joint is suspicious, and generating a rechecking record according to the rechecking result.
2. 2. The method of claim 1, wherein the industrial data link monitoring point obtaining time-amplitude raw data comprises: Receiving time-amplitude original data from industrial data chain monitoring points in real time, wherein each time-amplitude original data consists of a time stamp and corresponding signal amplitude, and the signal is derived from an industrial sensor; and periodically inquiring the sensor data, and storing the time stamp and the amplitude of the corresponding signal into the original data stream to form data points of the time stamp and the amplitude of the signal.
3. 3. An industrial data chain access method according to claim 2, wherein said constructing a two-dimensional coordinate system from time-amplitude raw data comprises: Constructing a two-dimensional coordinate system of an XY axis by using the obtained time-amplitude original data, wherein the X axis in the two-dimensional coordinate system of the XY axis represents a time stamp, and the Y axis represents signal amplitude; the data points of the time stamp and the signal amplitude are mapped into a two-dimensional coordinate system to form a set of points.
4. 4. A method of accessing an industrial data chain according to claim 3, wherein said outputting the time zone window data after setting the unit decision time slot parameter in the two-dimensional coordinate system and inputting the time slot sliding algorithm comprises: Constructing a time slot sliding algorithm, defining unit judgment time slot parameters in a two-dimensional coordinate system, generating time period window data, and sampling with a temporary length at the center of a current time period window to obtain a local segment; And calculating local fluctuation degree and event density according to the time period window data, and regulating the time period window length by the local fluctuation degree and the event density together to obtain the self-adaptive time slot.
5. 5. The method of claim 4, wherein the suppressing the anchor sequence by non-maximum suppression of the time period window data comprises: the adaptive time slot is used as the window length of a time period, the time range of an mth window is generated, window feature vectors are extracted on a local segment to form window data, anchor point scores are calculated for each window data, an initial anchor point is generated at the center of the window, non-maximum suppression is carried out on the initial anchor point, candidate anchor points in adjacent domains are merged by the suppression radius related to the window length, only the candidate anchor point with the highest score is reserved in the adjacent domains with the same suppression radius, and the candidate anchor point reserved after the non-maximum suppression forms a post-suppression anchor sequence.
6. 6. An industrial data on-chain access system for implementing the industrial data on-chain access method as claimed in any one of claims 1 to 5, comprising a raw data processing module, a high-association anchor point module, a cross-volume directory module and a visual view module; The original data processing module acquires time-amplitude original data from industrial data chain monitoring points, and constructs a two-dimensional coordinate system according to the time-amplitude original data; The high-association anchor point module inputs the inhibited anchor sequence into an association fragment type judgment algorithm to obtain high-association anchor point data, and the high-association anchor point data is subjected to time interval expansion, hash value extraction, time correction and aggregation index to obtain bookmark total data; The cross-volume catalog module performs head-to-tail butt joint verification by applying bookmark total table data to generate inter-volume connection record data, performs time sequence arrangement after performing missing segment supplementary recording on the inter-volume connection record data, and generates complete cross-volume catalog data; The visual view module constructs two visual views according to the complete cross-volume directory data, generates a continuity certificate based on the two views, then inputs a time interval threshold value to detect and obtain a gap mark, and rechecks the gap mark to obtain a rechecking result.

Description

Industrial data link access method and system Technical Field The invention relates to the field of industrial data processing, in particular to an industrial data on-link access method and system. Background The invention relates to an access method and system on an industrial data chain, which are suitable for collecting, storing and tracing equipment state monitoring data in the fields of intelligent manufacturing, energy chemical industry and the like. The existing industrial data chain access technology has obvious bottleneck, firstly, fixed time granularity anchoring data is adopted, transient anomalies such as short-time pressure spikes, vibration impacts and the like are easy to miss, a static threshold trigger mechanism is delayed in response to data characteristic changes, a complete risk evidence chain cannot be constructed, secondly, when historical data is split and archived, a roll boundary is often not aligned with an anchoring boundary on a chain, so that a cross-roll tracing broken chain is caused, continuity evidence is lacked in audit, and the prior art lacks an on-chain-under-chain cooperative mechanism, and is low in retrieval efficiency. Therefore, the invention provides an industrial data link access method and system. Disclosure of Invention In order to achieve the above purpose, the application provides an industrial data link access method and system, and the specific technical scheme is as follows: An industrial data link access method, comprising: S1, acquiring time-amplitude original data from industrial data chain monitoring points, and constructing a two-dimensional coordinate system according to the time-amplitude original data, wherein unit judgment time slot parameters are set in the two-dimensional coordinate system, time slot sliding algorithm is input into the time slot parameters, time slot window data are output, and the time slot window data are subjected to non-maximum suppression to obtain a suppressed anchor sequence; S2, inputting the suppressed anchor sequence into a relevant segment type judging algorithm to obtain high-relevance anchor point data, and obtaining a bookmark summary table through time interval expansion, hash value extraction, time correction and aggregation index; S3, performing end-to-end butt joint verification by applying bookmark total table data to generate inter-volume joint record data, and performing time sequence arrangement after performing missing segment supplementary recording on the inter-volume joint record data to generate complete cross-volume directory data; s4, constructing two visual views according to the complete cross-volume directory data, generating a continuity certificate based on the two views, inputting a time interval threshold value to detect and obtain a gap mark, and rechecking the gap mark to obtain a rechecking result. Preferably, the industrial data chain monitoring point acquiring time-amplitude raw data comprises the following steps: Receiving time-amplitude original data from industrial data chain monitoring points in real time, wherein each time-amplitude original data consists of a time stamp and corresponding signal amplitude, and the signal is derived from an industrial sensor; and periodically inquiring the sensor data, and storing the time stamp and the amplitude of the corresponding signal into the original data stream to form data points of the time stamp and the amplitude of the signal. Preferably, the constructing a two-dimensional coordinate system according to the time-amplitude raw data includes the following steps: Constructing a two-dimensional coordinate system of an XY axis by using the obtained time-amplitude original data, wherein the X axis in the two-dimensional coordinate system of the XY axis represents a time stamp, and the Y axis represents signal amplitude; the data points of the time stamp and the signal amplitude are mapped into a two-dimensional coordinate system to form a set of points. The step of outputting the window data of the time period after setting the unit judgment time slot parameter and inputting the time slot sliding algorithm in the two-dimensional coordinate system comprises the following steps: Constructing a time slot sliding algorithm, defining unit judgment time slot parameters in a two-dimensional coordinate system, generating time period window data, and sampling with a temporary length at the center of a current time period window to obtain a local segment; And calculating local fluctuation degree and event density according to the time period window data, and regulating the time period window length by the local fluctuation degree and the event density together to obtain the self-adaptive time slot. Preferably, the anchor sequence after the time period window data is suppressed by non-maximum suppression comprises the following steps: the adaptive time slot is used as the window length of a time period, the time range of an mth window is generated, window