Method for recommending operating commands for industrial plant
By receiving plant alarms and current status, comparing subsets that match the historical status list, running multiple simulations based on the simulation model, and recommending operation command variables with the highest quality value, it solves the difficulty of operators making appropriate operations in emergency situations. command issues to achieve safe and stable operation of the factory.
Patent Information
- Application Number
- CN202380046295.3
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Priority Date
- 2022-06-15
- Filing Date
- 2023-06-13
- Publication Date
- 2025-01-21
AI Technical Summary
In complex industrial factory environments, it is difficult for operators to quickly make appropriate operating commands when facing emergencies to effectively control the actuators of the factory.
By receiving the alarms and current status of the plant, comparing the subsets that match the historical status list, running multiple simulations based on the simulation model, each simulation based on the variables of the historical operating commands, and finally recommending the operating command variables with the highest quality value.
This method can provide operators with quick and effective operating command suggestions in emergency situations, helping the plant operate safely and stably.
Smart Images
Figure CN119343649A_ABST
Abstract
Description
Technical Field
[0001] The invention relates to the field of controlling industrial plants, in particular the field of controlling industrial plants by means of operating commands. The invention also relates to a program element, a computer-readable storage medium and a use. Background Art
[0002] Operating a plant is a very complex task, at least for some plants. A well-trained operator usually has a good understanding of the behavior of the plant, which enables the operator to execute operating commands that are well suited to control the plant in many situations. This may include emergency and / or alarm situations at the plant. However, it is suggested that, at least in some cases, the use of operating commands may be useful. Summary of the invention
[0003] The object of the invention is to provide a method for recommending an operating command. This object is achieved by the subject matter of the independent claims. Further embodiments are evident from the dependent claims and the following description.
[0004] One aspect relates to a computer-implemented method for recommending an operating command capable of controlling at least one actuator of an industrial plant, the method comprising the steps of: receiving an alarm from at least one sensor and / or from an operator of the plant, wherein the alarm is related to a current state of the plant;
[0005] Acquire a current state of the plant, the current state comprising at least one current process value and / or at least one current operation command related to the plant;
[0006] comparing a current state of the plant with a list of historical states of the plant, each historical state comprising a plurality of historical process values and / or at least one historical operation command associated with the plant;
[0007] If the current state matches at least a subset of the historical states, starting a simulation based on a simulation model of the plant and the matching historical states as a starting state;
[0008] Running multiple simulations, each simulation based on a variable of at least one historical operation command;
[0009] determining a quality value for each simulation in the plurality of simulations based on at least one quality criterion; and
[0010] The variant of the operating command is recommended so that the simulation has the highest quality value.
[0011] An industrial plant may be configured to run a process that may include multiple connectable sub-processes. The process may be a chemical process, a manufacturing process for any type of commodity, and / or other types of processes. The plant or process may be affected and / or controlled by an operating command, such as an operating command entered by an operator. The operating command is capable of controlling at least one actuator of the industrial plant. The actuators of the plant may include control elements, such as valves, heaters, motors, indicator elements, such as alarm indicators (e.g., lights and / or information), other elements. Sufficiently complex plants may contain multiple such elements. Controlling at least one actuator of the industrial plant may include direct control of the actuator and / or indirect control, such as by changing a set point.
[0012] The alarm may be caused and / or issued by at least one sensor and / or an operator of the plant. At least some of the sensors may be part of an actuator, such as a sensor indicating the position of an actuator, such as "valve half open". The alarm may be "related" to the current state of the plant, such as being caused and / or received within the same time frame as the current state and / or before that time frame.
[0013] An alarm may be or may indicate any critical situation of the plant. For example, if the criticality of the plant is represented between 0 and 9, with "0 = normal operation", or represented by "green / yellow / red", with "green = normal operation", then the critical situation may be "non-0" or "non-green". An alarm may be caused or sent by at least one sensor, or may be the result of merging and / or preprocessing process values from multiple sensors. Alternatively, an alarm may be received when an operator presses a predefined button. An alarm may be or may include alarms and / or notifications related to the plant and / or process.
[0014] When acquiring the current state of a plant, in many cases not all current process values of all sensors are available, but rather the current process values of a predefined subset of the sensors of the plant. The predefined subset of sensors may include sensors that are located outside the plant and are therefore "related" to the plant, such as a temperature sensor that is configured to measure the temperature of the environment outside the plant.
[0015] In order to compare the current state of the plant with the list of historical states of the plant, the historical states may be obtained and / or available in a log (e.g., a database) of the plant. The list of historical states may have a chronological order of the historical states. Each historical state may include a plurality of historical process values and / or at least one historical operation command associated with the plant. The historical state may have a number of process values and / or operation commands different from the comparable current state of the plant, such as more process values and / or operation commands. This is particularly true for large or complex plants, which may contain thousands of sensors providing process values. In one or more test runs, when the historical states are obtained, more powerful computers and / or larger memories may be used than during the production operation of the plant. The number of process current values and / or operation commands may vary from historical state to historical state. For example, the comparison step may compare a lower current process value and / or operation command with a higher number of historical process values and / or operation commands. At least in some cases, the comparison step may only consider a subset of historical process values and / or operation commands. Thus, a match to at least a subset of the historical states may be a match to all historical process values and / or operating commands, or a match to a true subset (i.e., a smaller value). Thus, a match may be a complete match or a partial match. A match may include an exact match, such as for sensors with binary and / or discrete values, or a "similarity" match, such as for a range of sensor values.
[0016] A match may result in zero clicks or one or more clicks. In the case of zero clicks, the simulation shall not be started. In the case of one hit, the simulation may start with all historical process values and / or operating commands contained in the matching historical state. In the case of two one or more clicks, the simulation may be started for each matching historical state. This may be a solution strategy for a practical problem for the operator: since various parameters may affect the operator's decision, there may still be a certain uncertainty as to whether the current situation is comparable to the situation that served as the basis for the selected strategy. In this method, the strategy may include considering all situations that may be related to the current state, as well as historical process values and / or operating commands, and recommending operating commands on this basis. The simulation method may include any simulation method applicable to the plant process.
[0017] Additionally or alternatively, historical data of the plant control system can be used to extract operational strategies. Historical data may include sensor information of processes and events generated by control system components; this allows capturing the plant state at different times. Furthermore, historical data may include so-called audit trails, i.e. events representing interactions between operators and the plant. Audit trails of operator commands and plant state information can be used as a data basis for strategy extraction.
[0018] Based on the matched historical state, as the starting state, multiple simulations can be run. As an option, the plant simulator can enter a state that is comparable to the historical interaction situation. As an option, the plant simulator can enter a state that is comparable to the situation of the plant operator. This initialization with the starting state provides an effective representation of the current state, allowing the simulation to test candidate strategies that may be recommended action commands. Each strategy can be formulated for a specific problem.
[0019] Multiple simulations can reflect or be based on strategies for handling the current state, in other words: testing possible reactions to alarms. One of the multiple simulations can use historical operating commands as data input. This can be used to test the effectiveness and startup state of the simulator. Other simulations may include changes in operating commands, such as changes in at least one historical operating command, such as by removing steps from historical commands, changing commands and / or other steps. Changes may include, for example, changes in the duration and / or time, intensity, etc. of the operating command. Activate actuators, different set point values for set point changes, and / or randomly remove steps. For example, a "workflow conditioner" module can automatically create variations (interferences) of the workflow, such as removing operator input / changing input values. Interfered workflows can be applied to process simulation instances. Each simulation can end based on an end criterion, such as "reaching a stable state" or "exceeding a predetermined number of simulation steps."
[0020] For each of the plurality of simulations, a quality value may be determined based on at least one quality criterion ("success criterion"). The quality criteria may include "reaching steady state", "did the condition of the plant deteriorate or improve at the end of the simulation?" "could the alarm be cleared?" comparison of a critical value with the value (as a result of historical operation commands), duration of the time period to clear the alarm, number of process steps to clear the alarm, return to normal state, alarm closure, process variable range to achieve ideal production, and / or further criteria. The criteria may be weighted and mapped to quality values, which may be numerical values for ease of comparison.
[0021] Based on the quality values, recommendations can be given, which variable led to the simulation run with the best results, more specifically: which operator action, actuator activation, etc. led to the best reaction for a given alarm. Furthermore, "counter-recommendations" can be given, for example based on simulations that led to undesirable plant behavior. The method can be advantageously used for training on real plant situations and / or as a basis for training artificial neural networks ANN for these situations.
[0022] In various embodiments, the method further comprises the following steps: acquiring a historical state list of the plant, each historical state comprising a plurality of historical process values and / or at least one historical operation command related to the plant; and storing the historical state list.
[0023] These steps may be performed in preparation for the above steps. Additionally or alternatively, the steps may be performed continuously during plant operation. The plant's historical status list may be stored in a repository, such as a database and / or in the cloud.
[0024] In various embodiments, at least a subset of the historical states is selected based on an algorithmic, heuristic, and / or ANN-based sensitivity analysis of a plurality of historical process values, and / or based on a random selection from a plurality of historical process values and / or at least one current process value.
[0025] Algorithmic sensitivity analysis can be based on, for example, variance decomposition, partial derivatives, or basis effects. Heuristic sensitivity analysis can be based on, for example, the expertise of experienced service personnel, which can be gathered, for example, in discussions and / or workshops. ANN-based sensitivity analysis can be based on multiple runs, for example, where the criterion or "reward" can be an alarm. In general, a sensitivity analysis may be based on the question "Which process values have the greatest impact on alarms or other events that are considered to be significant to the plant?" Sensitivity analysis and / or random selection can advantageously help reduce the number of values to be considered in the simulation. This may help reduce the complexity of the system so that useful statements can be made within a reasonable time frame.
[0026] In various embodiments, the comparison includes using a similarity function. The similarity function can be advantageously used to compare similar values. For example, for at least some sensors, a "similarity" range can be defined. The "similarity" range can be a kind of "value tolerance", for example, interpreting historical process values with a tolerance of 5%, so "90°C" can be interpreted as "between 85°C and 95°C".
[0027] In various embodiments, the matching of the current state with at least a subset of the historical states is based on a similarity analysis of elements of the current state with elements of the historical states. In particular, the similarity analysis may use a so-called Mueen algorithm for similarity search, as described at www.cs.unm.edu / ~mueen / FastestSimilaritySearch.html. Using similarity analysis may help further improve the importance, effectiveness, and / or relevance of the recommended variable combat orders.
[0028] In various embodiments, an alarm is at least one of the following: an emergency situation characterized by exceeding a predetermined threshold of at least one current process value, and / or a current operating command, an alarm of a plant alarm system, a trip of a predetermined actuator, and / or an activation of a subsystem. Even before a threshold is exceeded or an alarm is generated, a process variable may exhibit abnormal and chronically problematic behavior. A trip may include deactivating at least one device and / or interrupting an operating process.
[0029] In some embodiments, comparing the current state of the plant to the list of historical states includes comparing one state or a series of states. "One state" may be interpreted as comparing one current state to each historical state. "State sequence" may be interpreted as comparing a current state sequence (i.e., an ordered list of states, such as a chronological order) to a sequence of historical states. The sequence may include two, three, or more states, i.e., the same number of current states and historical states.
[0030] In various embodiments, each of the multiple simulations is based on a change in wait time, a change in set value range and / or a change in order, a random removal of steps. This may include providing the simulator with an operator action strategy and / or an operator workflow in a previously activated state, e.g. taking into account wait times, set value ranges, etc. as part of the strategy and recording the impact. Additionally or alternatively, the operator action strategy / operator workflow may be associated with a goal, e.g. alarm closure, efficiency represented by a process variable, etc. – which is provided to the simulator in the corresponding state and evaluating whether the effects displayed by the simulation are as expected.
[0031] In some embodiments, each simulation is terminated when one of the following criteria is met: when the simulated process value falls below a predetermined threshold; when a combined alarm of the plant alarm system is deactivated; and / or when a composite signal for normal plant operation is set.
[0032] The predefined thresholds may characterize a "normal" or non-critical state, i.e., the values of a subset of the relevant process variables do not exceed the thresholds and / or return to a normal state. A composite alarm may be an alarm that is set, reset, or deactivated by a plant alarm system, for example by aggregating simulated, current, or historical process values. A composite signal of normal plant operation may include values of a subset of the relevant process variables that are suitable for optimal plant operation.
[0033] In some embodiments, the quality criteria used to determine the quality value include at least one of the following: duration of the time period to clear the alarm, number of processing steps to clear the alarm, number of actuators involved and / or a predefined operating range for the process variable.
[0034] One aspect relates to a program element for executing the method described above and / or below when the program element is run on a processing unit and / or a control unit of a plant.
[0035] One aspect relates to a processing unit and / or a control unit of an industrial device, which is configured to carry out the method described above and / or below.
[0036] One aspect relates to a computer-readable storage medium, in which the above-mentioned computer program or computer program product is stored.
[0037] One aspect relates to the use of the above-mentioned and / or below-mentioned method and / or to a control unit configured to execute the method, recommend operating commands and / or control an industrial device by applying the recommended operating commands.
[0038] For further explanation, the present invention is described by the embodiments shown in the drawings. These embodiments are only to be regarded as examples and not as limitations. BRIEF DESCRIPTION OF THE DRAWINGS
[0039] Description of the drawings:
[0040] Figure 1 schematically shows an arrangement according to an embodiment;
[0041] Figure 2 is a flow chart according to an embodiment. DETAILED DESCRIPTION
[0042] Figure 1 The arrangement according to the embodiment is schematically shown. The arrangement includes an industrial plant 10, which can be configured for running a process. The process may include a plurality of connectable sub-processes. The process may be a chemical process, a manufacturing process of any type of commodity and / or a process of other types. The plant 10 or process may be affected and / or controlled by an operation command 22, such as input by an operator 20. The operation command can control at least one actuator 16 of the industrial plant. The actuator 16 of the plant 10 may include a control element, such as a valve, a heater, a motor, an indicator element, such as an alarm indicator (such as a light and / or a message), other elements. For the plant 10, a simulation model 18 can be obtained. The method for simulating 18 may include any simulation method applicable to the plant process. The process value associated with the plant 10 (e.g., a value that may affect the behavior of the plant from a sensor 14 inside or outside the plant 10) may have been monitored and stored as a historical process value 34. In addition, the operation command 22 may have been monitored and stored as a historical operation command 32. A plurality of historical process values 34 and / or historical operation commands 32 may be stored as a historical state list 30, wherein each entry of the list contains a historical process value 34 and / or a historical operation command 32 at a different time. The historical state list 30 may be in the form of a time series of historical states. The historical state list 30 may be stored in a repository, such as a database, which may be located in the cloud.
[0043] In addition, the current state 40 of the plant 10 is monitored, in particular continuously. The current state 40 includes at least one current process value and / or at least one current operating command 22 associated with the plant 10. When an emergency situation occurs, such as a combination of the current process value and / or the current operating command 22, the processing unit and / or the control unit 12 of the plant 10 can receive an alarm 45. The alarm 45 can be related to the current state 40 of the plant 10, for example, it can indicate an emergency situation. Then, the current state 40 of the plant 10 is obtained. The current state 40 can include at least one current process value and / or at least one current operating command 22 associated with the plant 10. The current state 40 can be compared with the historical state list 30. Based on the comparison result, one or more simulations can be run in the simulator 18. As a result of the simulation, the processing unit and / or the control unit 12 can output a recommendation for the operating command 50. Additionally or alternatively, the processing unit and / or the control unit 12 can output an operating command (not shown) based on the recommendation.
[0044] Figure 2 A flow chart 100 according to an embodiment is shown. In optional step 102, a list 30 of historical states of a plant 10 is obtained. Each entry of the list, i.e., each historical state 30, may include a plurality of historical process values 34 and / or at least one historical operation command 32 associated with the plant 10. In optional step 104, the list 30 of historical states is stored. In step 106, an alarm 45 is received from at least one sensor 14 and / or an operator 20 of the plant 10, wherein the alarm 45 is associated with a current state 40 of the plant 10. In step 108, the current state 40 of the plant 10 is obtained, the current state 40 including at least one current process value and / or at least one current operation command 22 associated with the plant 10. In step 110, the current state 40 of the plant 10 is compared with the list 30 of historical states of the plant 10, each historical state 30 including a plurality of historical process values 34 and / or at least one historical operation command 32 associated with the plant 10. If the current state 40 does not match a subset of at least one of the historical states 30, the method terminates in step 112.
[0045] If the current state 40 matches a subset of at least one of the historical states 30, then at step 114, a simulation is started based on the simulation model 18 of the plant 10 and the matched historical states 30 as a starting state. At step 116, a plurality of simulations are run, each simulation being based on a variable of at least one historical operating command 32. At step 118, for each of the plurality of simulations, a quality value is determined based on at least one quality criterion. At step 120, a variable of the operating command 32 that results in the simulation having the highest quality value is recommended.
[0046] Reference numerals list
[0047] 10Industrial Factory
[0048] 12 Processing unit and / or control unit
[0049] 14(s) sensors
[0050] 16(s) actuators
[0051] 18Simulation models and simulators
[0052] 20 Operators
[0053] 22 Operation Commands
[0054] 30 History Status
[0055] 32(multiple) historical operation commands
[0056] 34 Historical process values
[0057] 40 Current Status
[0058] 45 Alarm
[0059] 50 Operational commands and / or recommendations
[0060] 100 Flowchart
[0061] Steps 102 to 120
Claims
1. A computer-implemented method for recommending an operating command (32) capable of controlling at least one actuator (16) of an industrial plant (10), the method comprising the following steps: receiving an alarm (45) from at least one sensor (14) and / or from an operator (20) of the plant (10), wherein the alarm (45) is related to a current state (40) of the plant (10); Acquiring the current state (40) of the plant (10), the current state (40) comprising at least one current process value and / or at least one current operation command (22) related to the plant (10); comparing the current state (40) of the plant (10) with a list of historical states (30) of the plant (10), each historical state (30) comprising a plurality of historical process values (34) and / or at least one historical operation command (32) associated with the plant (10); If the current state (40) matches at least a subset of the historical states (30), starting a simulation based on a simulation model (18) of the plant (10) and the matching historical states (30) as a starting state; running a plurality of simulations, each simulation being based on a variable of at least one of the historical operation commands (32); for each simulation in the plurality of simulations, determining a quality value based on at least one quality criterion; as well as The variant of the operating command (32) that causes the simulation to have the highest quality value is recommended.
2. The method according to claim 1, further comprising the steps of: Acquire the list of historical states (30) of the plant (10), each historical state (30) comprising a plurality of historical process values (34) and / or at least one historical operation command (32) related to the plant (10); and The list of historical states (30) is stored.
3. The method according to any one of the preceding claims, wherein the subset of at least one of the historical states (30) is selected based on an algorithmic, heuristic and / or ANN-based sensitivity analysis of the plurality of historical process values (34) and / or based on a random selection from the plurality of historical process values (34) and / or the at least one current process value.
4. The method according to any one of the preceding claims, Wherein the comparing comprises using a similarity function.
5. The method according to any one of the preceding claims, The matching of the current state (40) with the subset of at least one of the historical states (30) is based on a similarity analysis of elements of the current state (40) and elements of the historical states (30).
6. The method according to any one of the preceding claims, The alarm (45) is at least one of the following: a critical situation characterized by exceeding a predefined threshold of at least one current process value and / or characterized by a current operating command (22); Alarms of the factory alarm system, Predetermined tripping of actuators, Subsystem startup, and / or Stopping of a subsystem.
7. The method according to any one of the preceding claims, Wherein comparing the current state (40) of the plant (10) with a list of historical states (30) includes comparing one state or a series of states.
8. The method according to any one of the preceding claims, Each simulation in the plurality of simulations is based on a variable of waiting time, a variable of a set value range and / or a random removal step.
9. The method according to any one of the preceding claims, Each simulation in the simulation is terminated when one of the following conditions is met: When the simulated process value is below a predefined threshold; When the combined alarm of the plant alarm system is deactivated; and / or When the combined signals are set for normal factory operation.
10. The method according to any one of the preceding claims, The quality criteria used to determine the quality value include at least one of the following: the duration of the time period during which the alarm (45) is released; the number of processing steps required to resolve the alarm (45), The number of actuators involved, and / or Predefined operating ranges for process variables.
11. A processing unit and / or a control unit (12) of an industrial plant (10), configured to perform the method according to any one of claims 1 to 10.
12. A program element for executing the method according to any one of claims 1 to 10 when the program element is run on a processing unit and / or a control unit (12) of an industrial plant (10).
13. A computer-readable storage medium, in which a computer program element according to claim 12 is stored.
14. Use of the method according to any one of claims 1 to 10, and / or use of a control unit configured to perform the method for recommending operating commands (32) and / or controlling an industrial plant (10) by applying the recommended operating commands (32).