CN-122018477-A - Method for monitoring operation state data of field device in production industry in real time

CN122018477ACN 122018477 ACN122018477 ACN 122018477ACN-122018477-A

Abstract

The invention relates to the technical field of data monitoring and discloses a method for monitoring operation state data of field equipment in production industry in real time, which comprises the steps of obtaining production level main data, constructing an area list table and an acquisition point list table, establishing a dynamic mapping mechanism to screen effective acquisition point sets, completing self-adaptive variable frequency acquisition and data quality verification, generating standard data messages and sending the standard data messages into a message queue; the method comprises the steps of storing effective information of a regional list table and an acquisition point list table into an edge side memory, distinguishing abnormal types, completing differentiated data restoration and full-quantity standardized processing, packaging the data into a processed data message, executing two-dimensional abnormal recognition combining a fixed absolute safety threshold and a working condition self-adaptive dynamic threshold, executing hierarchical control and root cause convergence on the recognized alarms, completing full-flow closed-loop binding from alarm triggering to treatment archiving, updating core configuration and processing rules of three steps before back feeding optimization, and constructing an equipment fault knowledge base by the acquisition point list table.

Inventors

Chen yaxing
HUANG JIACHEN
QIAN WANGPING
GU PENG
YANG FANGFANG

Assignees

江苏沙钢高科信息技术有限公司

Dates

Publication Date: 20260512
Application Date: 20260416

Claims (10)

1. The method for monitoring the operation state data of the field equipment in the production industry in real time is characterized by comprising the following steps of: constructing an area list table and an acquisition point list table from production level main data uniquely identified by a enterprise owner data platform synchronous belt, establishing a dynamic mapping mechanism to screen an effective acquisition point set, completing self-adaptive variable frequency acquisition and data quality verification according to equipment ID grouping, generating a standard data message and sending the standard data message into a message queue; Caching effective information of the area list table and the acquisition point list table into an edge side memory, continuously subscribing standard data messages in a message queue, storing in groups according to equipment ID and acquisition time stamp after analysis, executing two-stage space-time alignment of single equipment and cross equipment, distinguishing abnormal types, completing differentiated data restoration and full-quantity standardized processing, and packaging into processed data messages; Based on the processed data message, performing two-dimensional abnormality recognition by combining a fixed absolute safety threshold and a working condition self-adaptive dynamic threshold, synchronously completing trend abnormality judgment of a continuous acquisition period, performing hierarchical control and root cause convergence on the recognized alarm, and completing classified directional hierarchical data pushing; and automatically generating a standardized operation and maintenance work order, completing the whole-flow closed-loop binding from alarm triggering to disposal filing, synchronously updating an acquisition point list table to construct an equipment fault knowledge base based on core configuration and processing rules of three steps before the filing disposal record back feeding optimization, and periodically generating an equipment operation state analysis report.
2. The method for monitoring operation state data of field devices in production industry in real time according to claim 1, wherein the method for establishing a dynamic mapping mechanism is as follows: the method comprises the steps of synchronizing full-quantity real-time updated production level main data with unique marks from a business owner data platform, constructing a region list table and an acquisition point list table by combining the main data, wherein the region list table is newly added with a region working condition state, a region production effective period and a post authority matching field, and the acquisition point list table is newly added with an acquisition point semantic standardization, an acquisition point effective state, equipment failure mechanism association and a working condition adaptation label field; And establishing a real-time dynamic mapping matching mechanism, and simultaneously meeting the conditions that the area to which the acquisition point belongs is in the effective period of the production plan, the current login user has area access authority, the acquisition point is in an effective starting state under the current area working condition, and the acquisition point is included in an effective acquisition and pushing range, or else, is automatically shielded.
3. The method for monitoring the operation state data of the field equipment in the production industry in real time according to claim 1, wherein the method for collecting the self-adaptive variable frequency is as follows: the acquisition program reads an effective acquisition point set, establishes an acquisition channel according to equipment ID grouping and a field SMC system, and acquires real-time operation data, equipment real-time working conditions, production load and process set values of corresponding equipment and acquisition points; Setting basic sampling frequency of 1 time per second for key sampling points and basic sampling frequency of 1 time per 5 seconds for non-core auxiliary sampling points by combining equipment fault mechanism associated fields of the sampling points, dynamically adjusting the real-time sampling frequency by combining equipment real-time working condition fluctuation and data abnormality degree, synchronously completing data quality cross check and sensor state monitoring in each acquisition process, and finally generating standard data messages and sending the standard data messages into a message queue.
4. The method for monitoring operation state data of field devices in production industry in real time according to claim 1, wherein the two-stage space-time alignment method is as follows: When the background service is started, the effective information of the area list table and the collection point list table is cached to an edge side memory, a continuous subscription message queue receives standard data messages, and after the messages are analyzed, the messages are stored to an edge side local cache according to equipment ID and collection time stamp groups; And uniformly marking all the acquisition points of the same equipment in the same period as the reference time stamp of the time window by taking the starting time stamp of each acquisition period as a reference, completing the time stamp alignment of the acquisition points in single equipment, and uniformly aligning the cross-equipment time stamps of the equipment acquisition points associated with the upstream and downstream procedures of the whole production line by taking the uniform system time of a business owner data platform as a reference.
5. The method for monitoring operation state data of field devices in production industry in real time according to claim 1, wherein the method for repairing differential data is as follows: Aiming at the aligned time sequence data, selecting a reference acquisition point real-time effective value which has clear physical mechanism association with the acquisition point, calculating to obtain a repair value of the missing data by combining with a fixed mechanism association proportion determined by factory design parameters of equipment to complete the completion of the single acquisition point missing data in a time window; For abnormal data marked as a state to be checked, distinguishing abnormal types through mechanism association checking, performing interpolation restoration on false anomalies caused by instantaneous disturbance of a sensor, reserving original abnormal values for real anomalies caused by equipment operation state change and marking the real anomalies as abnormal characteristic data, performing full-scale standardization processing on the restored effective data, and finishing data aggregation processing on non-core acquisition point data under normal working conditions at the edge side.
6. The method for monitoring operation state data of field devices in production industry in real time according to claim 1, wherein the method for identifying the two-dimensional anomalies is as follows: setting a fixed absolute safety threshold interval comprising an absolute upper limit and an absolute lower limit for each acquisition point based on equipment factory design parameters, industry safety specifications and historical fault data, and triggering the highest-level emergency alarm when the real-time data exceeds the absolute safety threshold interval; and in the absolute safety threshold interval, setting a dynamically adjusted alarm threshold interval for each acquisition point, adjusting the alarm threshold range by combining the real-time production load of the equipment based on the theoretical parameter interval set by the current production process, and correcting the alarm threshold offset according to the full life cycle operation time of the equipment.
7. The method for monitoring operation state data of field devices in production industry in real time according to claim 1, wherein the method for judging trend anomalies is as follows: and aiming at the time sequence data of 5 continuous acquisition periods, judging whether the data change trend accords with the normal operation rule of the equipment, and directly triggering early abnormal early warning at the acquisition points which show continuous single upward or downward trend, do not exceed the dynamic alarm threshold interval and have trend slope exceeding the maximum range of the normal operation of the history.
8. The method for monitoring operation state data of field devices in production industry in real time according to claim 1, wherein the method for performing hierarchical control and root cause convergence of alarms is as follows: When a plurality of alarms occur in the same equipment, the alarm with the highest alarm level is taken as a root alarm, and the low-level alarms triggered by the other simultaneous periods are marked as derivative alarms; When the chain alarms of a plurality of devices occur in the same area, the highest-level alarm triggered first in time sequence is taken as a source alarm, the rest later-triggered alarms are marked as derivative alarms, and the derivative alarms are only recorded in the background and are not actively pushed to the front end.
9. The method for monitoring operation state data of field devices in production industry in real time according to claim 1, wherein the method for pushing directional hierarchical data is as follows: Establishing a Websocket long connection with the front end of the mobile terminal, completing identity verification and authority matching of a login user when connection is established, acquiring an effective area and an effective acquisition point set of a corresponding user, and establishing a mapping relation between user ID-Websocket connection-effective acquisition point set; Setting a self-adaptive heartbeat packet interval to maintain long connection stability, locally caching the equipment operation data and the alarm data in the disconnection period at the edge side when the network is disconnected, and pushing the data according to the priority after the network is recovered.
10. The method for monitoring the operation state data of field devices in the production industry in real time according to claim 1, wherein the method for binding the whole process in a closed loop is as follows: When the system triggers secondary and above alarms, a standardized operation and maintenance work order containing equipment ID, abnormal acquisition point information, alarm level, abnormal data, historical trend and associated fault mechanism information is automatically generated, and the standardized operation and maintenance work order is directionally distributed to equipment operation and maintenance personnel in a corresponding area according to the area post authority matching field; After the operation and maintenance personnel completes on-site disposal, the operation and maintenance personnel submits a disposal record containing the fault root cause, the disposal method, the disposal result and the equipment recovery state, and the system binds and files the disposal record and the original alarm event.

Description

Method for monitoring operation state data of field device in production industry in real time Technical Field The invention relates to the technical field of data monitoring, in particular to a method for monitoring operation state data of field equipment in the production industry in real time. Background In the field of monitoring the running state of equipment in the production industry, the prior art is based on a traditional manual spot inspection mode, an informationized mobile endpoint inspection monitoring scheme is derived, and the two schemes have systematic defects of suitability, instantaneity, accuracy and closed loop and cannot adapt to the fine management and control requirements of multi-branch-plant and multi-level production scenes; In the prior art, the acquisition configuration is fixed and stiff, the main data and the production organization are not adjusted synchronously, a static binding mode is adopted for a region and an acquisition point, invalid data redundancy and monitoring blind areas are easy to generate when the production line is changed and the equipment is overhauled, the fixed sampling frequency is adopted, the resource occupation and the fault characteristic capture cannot be considered, a sensor state verification mechanism is not arranged, the fault of the sensor and the abnormal operation of the equipment are easy to be confused, the problem of false alarm cannot be avoided in advance, the data preprocessing capability is weak, a cross-equipment time sequence alignment mechanism is not available, the time sequence dislocation of the data of a plurality of acquisition points easily causes the associated analysis failure, the processing mode of directly discarding the missing value and the abnormal value is adopted, the early fault characteristic of the equipment is easy to be lost, the total data processing mode introduces additional transmission and processing delay, the low-time delay monitoring requirement of an industrial field cannot be met, the abnormal recognition depends on a fixed threshold value, the fluctuation of the equipment aging dynamic adaptation cannot be realized along with the fluctuation, the fault condition of the equipment is impossible to be captured, the early warning alarm convergence mechanism is absent, the problem is easy to trigger the alarm wind alarm, the monitoring and the operation of the equipment is completely breaks down, the closed loop is not formed, and the self-optimizing and the long-term operation iterative effect cannot be realized; In view of this, there is a need to provide a method for real-time monitoring of production industry field device operational status data. Disclosure of Invention The invention aims to provide a method for monitoring field device running state data in real time in the production industry, which aims to solve the problems in the prior art, and is realized by the following technical scheme: in a first aspect, the method for monitoring the running state data of the field device in the production industry in real time provided by the embodiment of the invention specifically includes the following steps: step one, constructing an area list table and an acquisition point list table from production level main data uniquely identified by a enterprise owner data platform synchronous belt, establishing a dynamic mapping mechanism to screen an effective acquisition point set, completing self-adaptive variable frequency acquisition and data quality verification according to equipment ID grouping, generating a standard data message and sending the standard data message into a message queue; caching effective information of the area list table and the acquisition point list table into an edge side memory, continuously subscribing standard data messages in a message queue, storing in groups according to equipment ID and acquisition time stamps after analysis, executing two-stage space-time alignment of single equipment and cross equipment, distinguishing abnormal types, completing differentiated data restoration and full-quantity standardized processing, and packaging into processed data messages; Step three, based on the processed data message, executing the two-dimensional abnormality identification combining the fixed absolute safety threshold and the working condition self-adaptive dynamic threshold, synchronously completing the trend abnormality judgment of the continuous acquisition period, executing the hierarchical control and root cause convergence on the identified alarm, and completing the classified directional hierarchical data pushing; And step four, automatically generating a standardized operation and maintenance work order, completing the whole-flow closed-loop binding from alarm triggering to disposal filing, synchronously updating an acquisition point list table to construct an equipment fault knowledge base based on core configuration and processing rules of the three steps before the filing