CN-121997241-A - Integrated production operation platform based on digital intelligence twinning

CN121997241ACN 121997241 ACN121997241 ACN 121997241ACN-121997241-A

Abstract

The invention discloses an integrated production operation platform based on digital intelligent twin, which relates to the technical field of production operation, and aims to effectively improve the accuracy of abnormal identification and the timeliness of processing in complex production operation and enhance the stability and safety of the integral operation of a system by constructing a digital intelligent twin operation model of an operation node, integrating multi-source time sequence data, establishing a physical execution layer and a virtual mapping layer, introducing a state consistency constraint function to characterize synchronous deviation between the two layers, carrying out real-time iterative computation on a target operation execution model on the basis of the synchronous deviation to obtain a state deviation value of each operation node, checking the state consistency on the basis of signal time sequence logic and a robust online monitoring algorithm, generating an abnormal state judgment result, and further carrying out dynamic correction and closed-loop control of an operation process according to abnormal type execution instruction delay, path re-planning or state rollback.

Inventors

LI JINWEI
ZHOU HUA
DU HUA
LU HUA
CAI BIN
BAO WEIRONG
ZHANG LI
FENG ZHEN
ZHANG CHAO
HUANG PENG
LU GUANJUN
Cao Yuzhao
QIAN QIANG

Assignees

张家港保税区长江国际港务有限公司

Dates

Publication Date: 20260508
Application Date: 20260410

Claims (9)

1. The integrated production operation platform based on digital intelligence twinning is characterized by comprising the following components: The data acquisition module is used for acquiring initial process flow data and equipment state data of each operation unit in the target production operation system, constructing a corresponding PID (proportion integration differentiation) structured model and extracting state characteristic parameters of each node; The state mapping module is used for generating a multi-source time sequence data set by fusing the positioning data and the real-time sensing data based on the state characteristic parameters, and establishing a mapping relation set B= (B1, B2, & gt, bi, & gt, bn) of the state of the operation node, wherein bi represents the corresponding relation between the physical state and the virtual state of the ith operation node; The consistency constraint module is used for constructing a digital intelligent twin operation model according to the mapping relation set B, wherein the model comprises a physical execution layer and a virtual mapping layer, and a state consistency constraint function is introduced to represent synchronous deviation between the two layers; the execution control module is used for applying dynamic operation load and flow driving signals to the digital intelligent twin operation model based on an actual operation instruction and a scheduling strategy to obtain a target operation execution model; the deviation analysis module is used for carrying out real-time iterative computation on the target operation execution model by utilizing the multi-source time sequence data set, identifying the state deviation value of each operation node, checking according to the state consistency constraint function and generating an abnormal state judgment result; And the correction module is used for executing a correction strategy to the corresponding operation node according to the abnormal state judgment result, wherein the correction strategy comprises instruction delay, path re-planning or state rollback.
2. The integrated production operation platform based on digital intelligent twinning is characterized by comprising the steps of carrying out structural analysis on a process flow chart in a target production operation system, extracting pipeline connection relation, equipment type and process parameters, generating PID topology data, carrying out unique identification coding on each operation node based on the PID topology data, establishing an association mapping relation between the node and the equipment state data, obtaining operation state data of each equipment, carrying out timestamp alignment and data cleaning on the operation state data to form a state data set in a unified format, and extracting the state characteristic parameters representing the operation state of the node from the state data set according to preset feature extraction rules.
3. The integrated production operation platform based on the digital intelligent twin is characterized by constructing a digital intelligent twin operation model according to the mapping relation set B, and comprises the steps of respectively constructing a node state set of a physical execution layer and a simulation state set of a virtual mapping layer based on the corresponding relation between the physical states and the virtual states of all operation nodes in the mapping relation set B, establishing a node one-to-one mapping structure between the two layers according to PID topological data, carrying out unified dimension normalization processing on state vectors of all corresponding nodes on the basis of the node one-to-one mapping structure to form a comparable standard state vector pair, constructing a state consistency constraint function based on the standard state vector pair, embedding the state consistency constraint function into the digital intelligent twin operation model, and forming the digital intelligent twin operation model simultaneously comprising the physical execution layer, the virtual mapping layer and the state consistency constraint relation.
4. The integrated production operation platform based on the digital intelligent twin is characterized by applying dynamic operation load and flow driving signals to the digital intelligent twin operation model, and comprises the steps of obtaining actual operation instructions, analyzing and prioritizing the operation instructions according to a preset scheduling strategy to generate operation instruction sequences and corresponding time constraint parameters, mapping each operation instruction to a corresponding operation node in the digital intelligent twin operation model based on the operation instruction sequences and the time constraint parameters, distributing the dynamic operation load parameters to each operation node, and constructing the flow driving signals according to the connection relation among the operation nodes and the operation flow sequence.
5. The integrated production operation platform based on digital intelligent twinning as set forth in claim 4, wherein the dynamic operation load parameters comprise operation intensity, execution duration and resource occupation ratio.
6. The integrated production operation platform based on the digital intelligent twin system according to claim 4 is characterized in that flow driving signals are sequentially applied to each operation node according to time constraint parameters to drive state evolution of the digital intelligent twin operation model, and under the combined action of dynamic operation load and the flow driving signals, the digital intelligent twin operation model is subjected to state update and time sequence propulsion to generate a target operation execution model.
7. The integrated production operation platform based on the digital intelligence twinning is characterized in that real-time iterative computation is conducted on a target operation execution model by utilizing the multi-source time sequence data set, state deviation values of all operation nodes are identified, the method comprises the steps of conducting serialization processing on the multi-source time sequence data set according to a unified time step, conducting time alignment on the multi-source time sequence data set and node state vectors at corresponding moments in the target operation execution model to form a node comparison data sequence, constructing a difference computation model between a node state predicted value and an actual observed value based on the node comparison data sequence, conducting iterative computation on all operation nodes in a continuous time window to obtain node state deviation initial values, and conducting dynamic weighting smoothing processing on the node state deviation initial values according to historical operation data of all operation nodes to obtain the node state deviation values.
8. The integrated production operation platform based on the digital intelligent twin is characterized in that verification is carried out according to the state consistency constraint function to generate an abnormal state judgment result, and the integrated production operation platform based on the digital intelligent twin is composed of the steps of constructing a corresponding time sequence signal sequence based on the state consistency constraint function value of each operation node, building a signal time sequence logic judgment formula by combining preset operation flow constraint, state duration constraint and front-back node dependency constraint, inputting the time sequence signal sequence into a robust online monitoring algorithm, calculating the signal time sequence logic judgment formula at each moment to obtain a time sequence robustness value of the corresponding operation node, identifying whether the operation node has state mismatch, execution delay or flow sequence abnormality according to a comparison result of the time sequence robustness value and a preset abnormality judgment threshold, and carrying out association aggregation on abnormal identification results of each operation node according to an operation flow sequence to generate the abnormal state judgment result.
9. The integrated production work platform based on the digital intelligent twin system according to claim 8, wherein the instruction delay is executed preferentially for low serious anomalies, the path is re-planned for medium serious anomalies, and the state rollback is executed preferentially for high serious anomalies.

Description

Integrated production operation platform based on digital intelligence twinning Technical Field The invention relates to the technical field of production operation, in particular to an integrated production operation platform based on digital intelligence twinning. Background In port storage, transportation, loading and unloading production operation, multi-loop cooperative operation between ships, vehicles and storage tanks is involved, the prior art relies on a discrete information system and manual scheduling, digital twin and flow management are introduced, but the problem of timing deviation and semantic inconsistency between a fine-grained operation state and a virtual model still exists under complex working conditions, especially millisecond synchronization and consistency verification are difficult to realize in the fusion process of multi-source heterogeneous data (such as PID process data, equipment state data and positioning data), so that a 'state drift' phenomenon occurs in local operation nodes, namely the virtual model is updated and actual equipment does not respond or vice versa, thereby causing the safety risk of false triggering of instructions, path conflict or concealment. Disclosure of Invention The invention aims to provide an integrated production operation platform based on digital intelligence twinning, which aims to solve the defects in the background technology. In order to achieve the purpose, the invention provides the following technical scheme that the integrated production operation platform based on digital intelligence twinning comprises: The data acquisition module is used for acquiring initial process flow data and equipment state data of each operation unit in the target production operation system, constructing a corresponding PID (proportion integration differentiation) structured model and extracting state characteristic parameters of each node; The state mapping module is used for generating a multi-source time sequence data set by fusing the positioning data and the real-time sensing data based on the state characteristic parameters, and establishing a mapping relation set B= (B1, B2, & gt, bi, & gt, bn) of the state of the operation node, wherein bi represents the corresponding relation between the physical state and the virtual state of the ith operation node; The consistency constraint module is used for constructing a digital intelligent twin operation model according to the mapping relation set B, wherein the model comprises a physical execution layer and a virtual mapping layer, and a state consistency constraint function is introduced to represent synchronous deviation between the two layers; the execution control module is used for applying dynamic operation load and flow driving signals to the digital intelligent twin operation model based on an actual operation instruction and a scheduling strategy to obtain a target operation execution model; the deviation analysis module is used for carrying out real-time iterative computation on the target operation execution model by utilizing the multi-source time sequence data set, identifying the state deviation value of each operation node, checking according to the state consistency constraint function and generating an abnormal state judgment result; And the correction module is used for executing a correction strategy to the corresponding operation node according to the abnormal state judgment result, wherein the correction strategy comprises instruction delay, path re-planning or state rollback. Preferably, the step of extracting the state characteristic parameters of each node comprises the steps of carrying out structural analysis on a process flow chart in a target production operation system, extracting pipeline connection relation, equipment type and process parameters to generate PID topology data, carrying out unique identification coding on each operation node based on the PID topology data, establishing an association mapping relation between the node and the equipment state data, obtaining the running state data of each equipment, carrying out timestamp alignment and data cleaning processing on the running state data to form a state data set in a unified format, and extracting the state characteristic parameters representing the running state of the node from the state data set according to a preset characteristic extraction rule. The method comprises the steps of respectively constructing a node state set of a physical execution layer and a simulation state set of a virtual mapping layer based on the corresponding relation between physical states and virtual states of all operation nodes in the mapping relation set B, establishing a node one-to-one mapping structure between the two layers according to PID topological data, carrying out unified dimension normalization processing on state vectors of all corresponding nodes on the basis of the node one-to-one mapping structure to form comparable stan