Method for analyzing the impact of operator actions in industrial plants

By converting the topological process flow chart of the industrial processing system into a directed graph, constructing sub-graphs to show the interdependence between elements, the problem of difficult to predict and identify the interdependence of elements in industrial factories is solved, and the actions of operators are realized and a deeper understanding of process variables and safer process management.

CN113535640BActive Publication Date: 2025-02-18ABB (SCHWEIZ) AG
View PDF 1 Cites 0 Cited by

Patent Information

Application Number
CN202110403104.7
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Priority Date
2021-02-11
Filing Date
2021-04-14
Publication Date
2025-02-18
Estimated Expiration
2041-04-14

AI Technical Summary

Technical Problem

In industrial plants, it is difficult to predict and identify interdependence between elements, especially for infrequent actions, which may have unforeseen side effects.

Method used

By converting the topological process flow chart of the processing system into a directed graph, selecting the starting node and constructing a subgraph, the interdependence between multiple elements is demonstrated. This method can provide an approximate quantification of the effects of operator actions and possible side effects by generating a list of nodes and relationships.

Benefits of technology

Provides operators with insights on the effects and side effects of operator actions and process variables, helping operators better understand the effects of actions and prevent undesired side effects when necessary, thus providing safer process guidance.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN113535640B_ABST
    Figure CN113535640B_ABST
Patent Text Reader

Abstract

The present invention is entitled "Method for Analyzing the Impact of Operator Actions in an Industrial Plant". The present invention relates to the field of processing systems, in particular, processing systems for industrial plants. The present invention discloses a method for determining the interdependencies between multiple elements in an industrial processing system. The method comprises the following steps: providing a process flow diagram PFD of the topology of the processing system; converting the PFD into a directed graph, wherein each of the multiple elements is converted into a node, and each relationship between the multiple elements is converted into a directed edge; selecting a node from the multiple nodes as a starting node; and constructing a subgraph, wherein the subgraph consists of all nodes forward connected from the starting node, thereby showing at least one interdependency between the multiple elements in the subgraph.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The invention relates to the field of processing systems, in particular processing systems for industrial plants. The invention also relates to a computer program product, a computer readable storage medium and the use of the method. Background Art

[0002] In many cases, industrial plants are or contain a complex network of elements connected by flows of materials, energy and information. These connections make the elements interdependent and a change in one element, such as its value, may affect one or more of the other elements, possibly even distant elements. As a result, it may become difficult for plant operators, especially for actions that are not taken frequently, to predict the effects of changes in value and / or other "actions", including so-called "side effects". Furthermore, it may be difficult to identify whether atypical process values ​​have a relevant impact on the production process. Summary of the invention

[0003] Therefore, the object of the present invention is to provide a method for determining interdependencies between elements in a process system, in particular an industrial process system. This object is achieved by the subject matter of the independent claims. Further embodiments are evident from the dependent patent claims and the following description.

[0004] One aspect relates to a method for determining interdependencies between multiple elements in an industrial process system. The method comprises the following steps:

[0005] A process flow diagram (PFD) that provides the topology of the processing system;

[0006] converting the PFD into a directed graph, wherein each of the plurality of elements is converted into a node, and each relationship between the plurality of elements is converted into a directed edge;

[0007] Selecting a node from a plurality of nodes as a starting node; and

[0008] A subgraph is constructed, wherein the subgraph consists of all nodes forward connected from a starting node, thereby showing at least one interdependency between a plurality of elements in the subgraph.

[0009] A process system may be an industrial plant, such as that used in chemical and process engineering. A process system may be configured to produce and / or manufacture substances, such as materials and / or compounds. Examples of elements may include pumps, compressors, containers or pressure vessels, tanks, heat exchangers, furnaces, fans, cooling towers, valves, etc. A process flow diagram may be a piping and instrumentation diagram (P&ID) and / or another diagram that describes elements of a process system.

[0010] Converting the PFD into a directed graph can be done, for example, by an adjacency matrix that represents multiple elements as rows and columns, and the relationships between multiple elements as matrix entries, thereby representing directed edges between elements. Other representations of directed graphs can also be used. Once a directed graph is available, any of the multiple nodes can be selected as a starting node. In some cases, it may be useful to use an input element that can be triggered or otherwise changed by an operator as a starting node. In the sense of a node in the graph, this "input element" can be starting or stopping a motor, or can include changing a process variable. This operation can be done for one or for multiple selected starting nodes, for example, for process variables that are known to be highly correlated, or even for "atypical" process variables whose effects may not be completely clear. Then, a subgraph of forward-connected nodes can be constructed. Therefore, at least one interdependency is shown between multiple elements in the subgraph - that is, at least one interdependency can be observed. This part of the topological model represents a subgraph with all elements that may be affected.

[0011] The method advantageously provides the operator with insights into the effects, even so-called "side effects" of operator actions and / or process variables. A "side effect" may be a reaction by an element that the operator has not planned and / or foreseen. Side effects may include positive, negative, or negligible effects. Thus, operator actions and / or other changes may be subject to further study. In at least some cases, this may lead to a better understanding of the action, and / or in some cases, provide safer process guidance by preventing the operator from causing undesirable side effects.

[0012] In various embodiments, the method includes the further step of outputting all nodes and all relationships of the subgraph for the subgraph. The output may include generating a list of all nodes representing devices and assets that may be affected by the operator action, i.e., devices and assets that are related to a change in the start node or its value. The output may include providing an approximation or "verbally" quantification of the effects (including side effects) of the operator action. The output may include visualizing the cause-effect relationships within the process plant for engineers to verify, for example as a dynamically created process graph or cause-effect matrix.

[0013] In various embodiments, the method comprises the following further step: Searching for cyclic interdependencies in the subgraph. One idea could be to perform a search in the subgraph with directional flows of material, energy and / or information flows, thereby creating a list of elements that may be affected by an operator action. The list of elements with links known from the topology defines a subset of the topological graph, the subgraph, and will be searched for cycles. The list of potentially affected elements and the detected cycles will be presented to the operator. This can facilitate the identification of possible or real cycles that have a self-reinforcing effect in the production process, for example. In a recycle loop. The one or more cycles may occur as a reaction to an operator action.

[0014] In various embodiments, the relationships between multiple elements include material flows, energy flows, and / or information flows. This reflects or addresses the fact that an industrial plant is a network, in some cases a complex network, of elements connected by flows of materials, energy, and information. Thus, relationships involve different kinds of connections that make elements interdependent and that make a change in one element have an effect on one or more other elements.

[0015] In various embodiments, the directed edges are attributed edges. An attributed edge may be associated with a quantitative influence on another element. This may facilitate a quantitative assessment of the effect(s) of an operator action and / or a change in a process variable. In cases where a quantitative effect of the flow is not directly available from the PFD, the PFD may be combined with, for example, measurement data and state estimates of the flow under predetermined conditions. Again, historical or simulation data may be used for this and / or to derive causal relationships. This may also be the basis for "linguistically" quantifying the (side effects) of an operator action.

[0016] In some embodiments, at least one edge of the subset of attribution edges has a fuzzy attribute. Examples of fuzzy attributes are fuzzy variables and / or fuzzy logic rules. Fuzzy attributes may include rules that describe effects between elements. For example, a rule may describe:

[0017] If the liquid level in reactor A is above 50%, the flow in pipe B starts.

[0018] If there is flow in pipe B, the liquid level in storage tank C will rise.

[0019] These rules can be attributed to human experts or can be derived from historical data.

[0020] In some embodiments, the start node is a node with only outgoing edges. This is a node that no edges point to. In other words, the start node is a "source-only node". This may be related to "actual input", i.e., the change is independent of other elements of the process or plant.

[0021] In some embodiments, the process flow diagram is a piping and instrumentation diagram (P&ID diagram). Advantageously, this provides a common format for describing a process or plant.

[0022] One aspect relates to a computer program product comprising instructions which, when executed by a computer, cause the computer to perform the method described above and / or below.

[0023] One aspect relates to a computer-readable storage medium on which is stored a computer program or a computer program product as described above.

[0024] One aspect relates to the use of a method as described above and / or below for searching for cyclic interdependencies in a process flow diagram (PFD) of a processing system.

[0025] One aspect relates to an industrial plant or processing system comprising a computer having stored thereon instructions which, when executed by the computer, cause the computer and / or the industrial plant to perform the methods described above and / or below.

[0026] For further clarification, embodiments of the present invention are described with the aid of the embodiments shown in the accompanying drawings. These embodiments are to be considered as examples only and not restrictive.

[0027] Description of the drawings

[0028] Figure 1 Schematically illustrates an example of analyzing the effect of an operator's actions according to an embodiment;

[0029] Figure 2 A flowchart for identifying affected elements and finding loops according to an embodiment is schematically shown.

[0030] Figure 3 A flow chart of a method according to an embodiment is schematically shown. DETAILED DESCRIPTION

[0031] Figure 1An example 100 of analyzing the effects of operator actions according to an embodiment is schematically shown. Operator actions 110 can trigger a graph analyzer 120 to evaluate a directed graph 140. The directed graph 140 can use a process flow diagram (PFD) - examples of elements shown include tanks and level measurement devices - and convert the PFD into a directed graph 140, which can describe these elements in an embodiment. Starting from a starting node in the directed graph 140, the directed graph 140 is traversed. By traversing from the starting node to its neighbors and its neighbors' neighbors (and so on), a subgraph 150 is created. The subgraph 150 contains all elements affected by the operator's actions. In a simple implementation of the method, the subgraph can be presented to the operator. Additionally or as an alternative, a subgraph can be searched for loops in a loop analyzer 160, that is, a path within the graph that returns to a previous element or first element in the path. For example, the loop analyzer 160 can use Johnson's algorithm to find the shortest path between all pairs of vertices in a directed graph. Additionally or alternatively, a list of all elements and all relations of the subgraph is output, for example in the display 180 .

[0032] Figure 2 Schematically illustrates a flow chart 200 for identifying affected elements and finding loops according to an embodiment. In step 201, an operator action is performed and / or a process value is changed. The process value may be selected by the operator to investigate the effect. The operator action and / or process value linked to the element. In the directed graph 140 (see Figure 1 ), this element is used as a starting point for a breadth-first search or alternatively for a depth-first search, which may not explore all paths, like Dijkstra's shortest path algorithm. In step 202, the search creates a list of affected elements, which creates a subgraph 150. Subgraph 150 also includes edges defined by energy, material, and information flows. In step 203, subgraph 150 can add the nodes and edges it traverses. In step 204, the subgraph defined by the nodes and edges is searched for a loop, which may introduce unexpected self-reinforcing responses to operator actions or atypical process values. In step 205, the list is presented to the operator.

[0033] Figure 3A flowchart 300 of a method for determining interdependencies between multiple elements in a processing system according to an embodiment is schematically shown. In step 301, a process flow diagram PFD of the topology of the processing system is provided. In step 302, the PFD is transformed into a directed graph, wherein each of the multiple elements is transformed into a node, and each relationship between the multiple elements is transformed into a directed edge. In step 303, a node in the multiple nodes is selected as a starting node. In step 304, a subgraph is constructed, wherein the subgraph consists of all nodes forward connected from the starting node, thereby showing at least one interdependency between the multiple elements in the subgraph. In optional step 305, a list of all elements and all relationships of the subgraph is output.

Claims

1. A computer-implemented method for determining interdependencies between a plurality of elements in an industrial process system, the method comprising the steps of: Provide a process flow diagram of the topology of the processing system; converting the process flow graph into a directed graph, wherein each of the plurality of elements is converted into a node, and each relationship between the plurality of elements is converted into a directed edge, wherein the relationship between the plurality of elements represents a material flow, an energy flow, and / or an information flow; Select one of the multiple nodes as the starting node; constructing a subgraph, wherein the subgraph consists of all nodes forward-connected from the starting node, thereby illustrating at least one interdependency between the plurality of elements in the subgraph; For the subgraph, output all nodes and all relationships of the subgraph; searching for cyclic interdependencies in the subgraph; as well as The affected elements and detected cycles are output to identify self-reinforcing effects in the production process as a self-reinforcing response to operator actions.

2. The method according to claim 1, in, The directed edge is an attributed edge.

3. The method according to claim 2, in, At least one edge of the subset of attributed edges has a fuzzy attribute.

4. The method according to any one of the preceding claims 1 to 2, in, The start node is a node that has only outgoing edges.

5. The method according to any one of the preceding claims 1 to 2, in, The process flow diagram is a piping and instrumentation diagram (P&ID) diagram.

6. A computer program product comprising instructions which, when said program is executed by a computer, cause said computer to perform the method according to any of the preceding claims 1 to 5.

7. A computer-readable storage medium having stored thereon a computer program comprising instructions which, when executed by a computer, cause the computer to perform the method according to any one of the preceding claims 1 to 5.

8. Use of the method according to any one of claims 1 to 5 for searching for cyclic interdependencies in a process flow diagram of a processing system.

9. An industrial plant or processing system comprising a computer having stored thereon a computer program comprising instructions which, when executed by the computer, cause the computer, the industrial plant and / or the processing system to perform a method according to any one of claims 1 to 5.

Citation Information

Patent Citations

  • Large-system graph theory decomposition method based on Pearson correlation

    CN108762201A