Plant operation support equipment

Abstract

To provide a plant operation support device capable of improving the accuracy in manual intervention regardless of experiences and skills of an operator.SOLUTION: A plant operation support device includes a prediction processing part which predicts a change in a process value of a process of a plant by use of a process model corresponding to the process. Upon manual intervention executed by an operator of the plant after prediction that the process value exceeds a preset threshold value, the prediction processing part predicts a change in the process value obtained by taking the executed manual intervention into account, assumes manual intervention different from the executed manual intervention, and also predicts a change in the process value obtained by taking the assumed manual intervention into account. The plant operation support device further includes a display processing part which displays a first trend graph of the process value obtained by taking the executed manual intervention into account superimposed on a second trend graph of the process value obtained by taking the assumed manual intervention into account.SELECTED DRAWING: Figure 1

Description

[Technical Field]

[0001] The present disclosure relates to a plant operation support device that supports the operation of a plant. [Background technology]

[0002] The operation support device disclosed in Patent Document 1 below is equipped with a parameter correction means that receives plant data obtained from a plant such as a thermal power plant and simulator data calculated using a plant simulator as inputs, and corrects the value of a control parameter using the deviation between a predetermined process value and its target value as an index.

[0003] A dynamic simulator, which is one type of plant simulator, has the function of predicting the dynamic behavior of process values ​​using a process model. Operation support devices perform feedback control (e.g., PID control) of control parameters so that the process values ​​reach their target values.

[0004] However, even when the above-mentioned feedback control is performed, there are cases where the predicted process value is predicted to exceed a predetermined allowable value. In this case, an alarm may be generated to prompt the plant operator to take manual control action. Manual control is so-called manual intervention in which the operator manually manipulates the control parameters. Manual intervention can be performed using the operation and monitoring means disclosed in Patent Document 2 listed below. After performing manual intervention, the operator confirms that the predicted process value does not exceed the allowable value. [Prior art documents] [Patent documents]

[0005] [Patent Document 1] Patent No. 4546332 [Patent Document 2] Japanese Patent Application Laid-Open No. 2002-92584 Summary of the Invention [Problem to be solved by the invention]

[0006] However, because manual intervention depends heavily on the experience and skill of the operator, the details of the operation, such as the amount and timing of control parameter manipulation, are not always appropriate. Therefore, it is desirable to improve the accuracy of manual intervention regardless of the operator's experience or skill by evaluating the appropriateness of the operation and reflecting it in the next manual intervention.

[0007] Therefore, an object of the present disclosure is to provide a plant operation support device that can improve the accuracy of manual intervention regardless of the experience or skill of the operator. [Means for solving the problem]

[0008] A first aspect of the present disclosure relates to a plant operation support device that supports plant operation. The plant operation support device includes a prediction processing unit that predicts a change in a process value of the plant using a process model corresponding to the process. When a manual intervention is performed by an operator of the plant after the process value is predicted to exceed a predetermined threshold, the prediction processing unit predicts a change in the process value taking into account the performed manual intervention, and also predicts a change in the process value taking into account a manual intervention different from the performed manual intervention. The plant operation support device further includes a display processing unit that superimposes a first trend graph of the process value taking into account the performed manual intervention and a second trend graph of the process value taking into account the assumed manual intervention.

[0009] The second aspect has the following feature in addition to the first aspect: the display processing unit displays the content of the performed manual intervention and the content of the assumed manual intervention in a comparative manner on a screen different from a trend screen on which the first trend graph and the second trend graph are superimposed.

[0010] A third aspect has the following characteristics in addition to the first or second aspect: the prediction processing unit predicts a start point at which the process value exceeds the threshold, and then predicts an end point at which the process value does not exceed the threshold by performing the manual intervention; and the display processing unit superimposes and displays the first trend graph and the second trend graph in the section from the start point to the end point. [Effects of the Invention]

[0011] According to the present disclosure, by superimposing a first trend graph of an actually performed manual intervention and a second trend graph of a planned manual intervention, the appropriateness of the performed manual intervention can be evaluated. The operator can reflect the evaluation results of the manual intervention in the next manual intervention. Therefore, the accuracy of the manual intervention can be improved regardless of the operator's experience or skill. [Brief explanation of the drawings]

[0012] [Figure 1] 1 is a block diagram showing a configuration example of a plant operation support device according to an embodiment; [Figure 2] FIG. 1 illustrates a manual intervention performed by a plant operator. [Figure 3] FIG. 10 is a diagram illustrating the evaluation of the adequacy of manual intervention. [Figure 4] FIG. 2 is a conceptual diagram illustrating an example of the hardware configuration of a processing circuit included in the plant operation support device. DETAILED DESCRIPTION OF THE INVENTION

[0013] Hereinafter, embodiments will be described with reference to the drawings. Common or corresponding elements in the various drawings will be denoted by the same reference numerals, and descriptions thereof will be simplified or omitted.

[0014] FIG. 1 is a block diagram showing an example of the configuration of a plant operation support device 1 according to an embodiment.

[0015] The plant operation support device 1 supports the operation of a plant (not shown). The plant is, for example, a chemical plant that is capable of predicting the dynamic behavior of process values ​​using a known process model. Examples of chemical plants include ethylene production plants, oil refinery plants, polymer production plants, and beer production plants.

[0016] The plant operation supporting device 1 includes an actual process information collecting unit 2, a control device 3, a sensor 4, an HMI 5, a display processing unit 6, a prediction processing unit 7, and a process prediction information collecting unit 8.

[0017] The actual process information collection unit 2 has a process actual measurement value database 21 and an operation information database 22. The actual process information collection unit 2 acquires actual measurement values ​​of process values ​​and control values ​​and control parameters that the control devices 3 output to the controlled object from control devices 3, which are typified by programmable logic controllers (PLCs) and distributed control systems (DCSs), and stores the acquired process actual measurement values, control values ​​and control parameters in the process actual measurement value database 21. The control devices 3 acquire actual measurement values ​​of process values ​​measured by sensors 4 installed in the plant.

[0018] In addition, when a plant operator manually intervenes in the control device 3 via the HMI 5, the actual process information collection unit 2 collects the operation details of the manual intervention (including the operation time) and stores them in the operation information database 22.

[0019] The display processing unit 6 can display trend graphs on the HMI 5 so that an operator can check time-series changes in process actual measured values, control values, and control parameters. The HMI 5 has a trend screen S1 and a history screen S2, which will be described later.

[0020] The prediction processing unit 7 acquires process actual values ​​from the process actual value database 21 and uses the acquired process actual values ​​to execute a prediction function 72 based on a process model 71. The process model 71 is a well-known model used to predict the dynamic behavior of process values. The process value may be, for example, the raw material inlet temperature of a reactor in a chemical plant. Examples of raw material inlet temperatures include the raw material inlet temperature of a pyrolysis tube in an ethylene production plant, the raw material inlet temperature of a distillation tower in an oil refinery, the inlet temperature of steam used to heat a reactor in a polymer production plant, and the inlet temperature of steam for a feed water tank in a beer production plant. The control values ​​may be, for example, the openings of various valves in a plant and the output of heaters, etc. The prediction function 72 determines the time at which the predicted value EV exceeds a predetermined lower limit LM as the predicted alarm occurrence time ta. The predicted alarm occurrence time ta is stored in the process predicted value database 81.

[0021] Furthermore, if an actual operator performs manual intervention (see "MV+1.0%" in FIG. 3) after a predictive alarm is issued, the prediction processing unit 7 predicts a pattern of the predicted value EV that takes into account the performed manual intervention. The predicted pattern of the predicted value EV is stored in the supposed operation information database 82 together with the details of the manual intervention (including the implementation times "11:08, 09, 10, 11" in FIG. 3). Here, manual intervention includes changes to the control value and control parameters such as the control gain and time constant.

[0022] At this time, the prediction processing unit 7 assumes that a manual intervention (see "MV+2.0" in FIG. 3) different from the manual intervention actually performed (MV+1.0%) will be performed. The assumed manual intervention can be determined, for example, by multiplying the performed manual intervention by a predetermined coefficient (e.g., 200%, 50%, etc.). The prediction processing unit 7 further predicts a pattern of the predicted value EVs that takes the assumed manual intervention into account. The predicted pattern of the predicted value EVs is stored in the assumed operation information database 82 together with the content of the manual intervention (including the implementation time "11:08, 09, 10, 11" shown in FIG. 3). Note that a plurality of different manual interventions may be assumed depending on the performance of the processing circuit 10, which will be described later. For example, two types of manual interventions (MV+2.0, MV+0.5) may be assumed, and the patterns of the predicted value EVs may be predicted respectively.

[0023] The process prediction information collection unit 8 has a process predicted value database 81 and an expected operation information database 82. The process prediction information collection unit 8 collects the process predicted values ​​predicted by the prediction processing unit 7, the control values ​​and control parameters used for the prediction, and stores them in the process predicted value database 81. The expected operation information database 82 stores the details of performed manual interventions, the details of expected manual interventions, the patterns of predicted predicted values ​​EV, and the patterns of predicted predicted values ​​EVs.

[0024] Next, the operation of the plant operation supporting device 1 will be described taking as an example a case where the process value is controlled to the set value SV.

[0025] Fig. 2 is a schematic diagram illustrating manual intervention performed by a plant operator. Fig. 2 is a graph showing the transition of a process value together with the occurrence time ta of a predicted alarm and the recovery time tr. In Fig. 2, PV, shown by a solid line, is the actual measured value of the process value, and EV, shown by a two-dot chain line, is the predicted value of the process value. Furthermore, LM is the lower limit of the process value that is allowed for quality control of the plant.

[0026] As shown in Figure 2, the actual measurement value PV transitions from the control target value SV toward the lower limit LM, and at time ta, when it is predicted that the predicted value EV will fall below the lower limit LM at time tb, a predictive alarm is issued on the HMI 5. Upon confirming the predictive alarm, an operator performs manual intervention by manually manipulating the control value, control parameters, etc. so that the predicted value EV does not reach the lower limit LM. In this embodiment, the control value, control parameters, etc. are the opening of the valve MV provided in the reaction tank, and manual intervention is described using an example in which the opening of the valve MV is increased by +1.0%.

[0027] After manual intervention, at time tr when the operator confirms that the predicted value EV does not fall below the lower limit LM, the HMI 5 returns from the predicted alarm to the normal state. As mentioned above, manual intervention is largely dependent on the operator's experience and skill, so it is desirable to evaluate the appropriateness of the operation and reflect this in the next manual intervention.

[0028] The plant operation support device 1 has a function of evaluating the validity of the manual intervention after recovery from a predictive alarm through manual intervention, i.e., after the predicted alarm recovery time tr has elapsed. Fig. 3 is a diagram for explaining the evaluation of the validity of the manual intervention.

[0029] First, the prediction processing unit 7 executes a prediction function 72 based on a process model 71, using the process values, control values, and control parameters acquired from the process actual measurement value database 21. The predicted value EV of the process value obtained in this way and the control values ​​and control parameters used for the prediction are stored in a process predicted value database 81 of the process prediction information collection unit 8. The prediction function 72 stores the time at which the predicted value EV of the process value to be managed is predicted to fall below a preset lower limit value LM in the process predicted value database 81 as a predicted alarm occurrence time ta.

[0030] If an operator performs manual intervention (MV+1.0%) after the predicted alarm occurrence time ta, the prediction processing unit 7 predicts a pattern of the predicted value EV taking into account the performed manual intervention, and stores the predicted pattern of the predicted value EV together with the details of the manual intervention (including the implementation times "11:08, 09, 10, 11") in the supposed operation information database 82.

[0031] In this case, the prediction processing unit 7 assumes that a manual intervention (MV+2.0) different from the actual manual intervention (MV+1.0%) will be performed, predicts a pattern of the predicted value EVs taking this assumed manual intervention into account, and further stores the predicted pattern of the predicted value EVs together with the details of the manual intervention (including the implementation times "11:08, 09, 10, 11") in the assumed operation information database 82. Note that multiple different manual interventions may be assumed depending on the performance of the processing circuit 10 described below. For example, it is possible to assume two-valued manual interventions (MV+2.0, MV+0.5) and predict the patterns of the predicted value EVs respectively. This improves the accuracy of the validity evaluation. Furthermore, different types of control values ​​or control parameters may be selectable for multiple assumed manual interventions. For example, different valve openings, gains, or time constants may be selectable, or multiple combinations may be selectable.

[0032] Furthermore, the prediction processing unit 7 stores in the process predicted value database 81 the time at which it predicts that the predicted value EV will not fall below the lower limit value LM after the time ta as the predicted alarm recovery time tr.

[0033] The operator can display the actual process measurement value PV as a trend graph on the HMI 5 to check the time series changes.

[0034] In the initial display state, a trend graph is displayed based on the process actual measured values ​​PV stored in the actual process information collecting unit 2. At this time, the predicted alarm occurrence section Op from the predicted alarm occurrence time ta to the predicted alarm recovery time tr is displayed in a box. When the operator selects a predicted alarm occurrence section Op, for example, by operating the cursor, the actual measured values ​​PV in the predicted alarm occurrence section Op may be highlighted. Note that multiple predicted alarm occurrence sections Op may be displayed. In this case, allowing the operator to select one predicted alarm occurrence section Op enables manual intervention and evaluation in each predicted alarm occurrence section Op.

[0035] The display processing unit 6 displays a first trend graph Gf1 of actual measured values ​​PV and a second trend graph Gf2 of predicted values ​​EVs superimposed on the trend screen S1 shown in Fig. 3 for the selected predicted alarm occurrence section Op. At this time, another trend graph Gf3 for the case where manual intervention is not performed may also be displayed. Circles and triangles on the first trend graph Gf1 and second trend graph Gf2 indicate the timing of manual intervention.

[0036] The manual intervention timing can also be indicated by other display methods (for example, by displaying it in red). The two trend graphs Gf1 and Gf2 have different manual intervention timings, but they may be the same timing. This allows for accurate evaluation of the manual intervention timing. However, by reducing the number of manual intervention timings for the predicted value EVs, the amount of data to be stored can be reduced.

[0037] 3, the history of actual manual interventions and the history of assumed manual interventions are displayed in a comparative manner on a history screen S2, which is different from the trend screen S1 that displays the trend graphs Gf1 and Gf2 superimposed on each other. This allows the operator to easily associate the details of the manual interventions with the trend graphs Gf1 and Gf2.

[0038] As described above, according to this embodiment, the validity of the performed manual intervention (MV+1.0%) can be evaluated by superimposing the first trend graph of the actual manual intervention (MV+1.0%) and the second trend graph of the assumed manual intervention (MV+2.0%). In the example shown in FIG. 3, it can be seen that the assumed manual intervention (MV+2.0%) leads to more stable plant operation than the performed manual intervention (MV+1.0%). The operator can reflect the evaluation results of the manual intervention in the next manual intervention. Therefore, it is possible to improve the accuracy of the manual intervention regardless of the operator's experience or skill.

[0039] Furthermore, by displaying the history of manual interventions in addition to the superimposed display of trend graphs Gf1 and Gf2, the operator can easily correlate the details of manual interventions with the trend graphs Gf1 and Gf2, thereby enabling the evaluation of manual interventions to be performed in a short time.

[0040] Furthermore, by highlighting the selected predicted alarm occurrence section Op, it is possible to clarify the location where manual intervention should be evaluated, which is particularly advantageous when a predicted alarm occurrence section Op is present.

[0041] The specific structure of the plant operation support device 1 is not limited, and may be, for example, as follows. FIG. 4 is a diagram showing an example of the hardware configuration of the processing circuit 10 included in the plant operation support device 1. The functions of the tension control device 11 can be realized by the processing circuit 10 shown in FIG. 4. The processing circuit 10 may be dedicated hardware 10a. The processing circuit 10 may include a processor 10b and a memory 10c. The processing circuit 10 may be partially formed as dedicated hardware 10a and further include a processor 10b and a memory 10c. In the example of FIG. 4, the processing circuit 10 is partially formed as dedicated hardware 10a, and the processing circuit 10 also includes a processor 10b and a memory 10c. The memory 10c may also serve as each of the databases 21, 22, 81, and 82.

[0042] At least a portion of the processing circuitry 10 may be at least one dedicated hardware 10a, which may be, for example, a single circuit, a composite circuit, a programmed processor, a parallel programmed processor, an ASIC, an FPGA, or a combination thereof.

[0043] The processing circuit 10 may include at least one processor 10b and at least one memory 10c. In this case, each function of the tension control device 11 is realized by software, firmware, or a combination of software and firmware. The software and firmware are written as programs and stored in the memory 10c. The processor 10b realizes the functions of each unit by reading and executing the programs stored in the memory 10c.

[0044] The processor 10b is also called a CPU (Central Processing Unit), central processing unit, processing device, arithmetic unit, microprocessor, microcomputer, or DSP. The memory 10c is, for example, a non-volatile or volatile semiconductor memory such as RAM, ROM, flash memory, EPROM, or EEPROM. In this way, the processing circuit 10 can realize each function of the tension control device 11 by hardware, software, firmware, or a combination of these.

[0045] Although the embodiments of the present invention have been described above, the present invention is not limited to the above embodiments and can be implemented in various modifications without departing from the spirit of the present invention. The plant is not limited to a chemical plant, and the present invention can be applied to any plant in which the dynamic behavior of a process value can be predicted using a process model.

[0046] In the above embodiment, the appropriateness of manual intervention is evaluated after recovery from a predictive alarm, but the appropriateness can be evaluated after manual intervention has been performed, even before recovery from a predictive alarm. This allows the evaluation results to be quickly reflected in manual intervention, which is particularly advantageous when the number of manual interventions is relatively high.

[0047] Furthermore, when the number, quantity, amount, range, etc. of each element is mentioned in the above embodiments, the present invention is not limited to the mentioned numbers unless otherwise specified or clearly specified in principle. Furthermore, the structures, etc. described in the above embodiments are not necessarily essential to the present invention unless otherwise specified or clearly specified in principle. [Explanation of symbols]

[0048] 1... Plant operation support device, 6... Display processing unit, 7... Prediction processing unit, 71... Process model, Gf1... First trend graph, Gf2... Third trend graph, S1... Trend screen, S2... History screen

Claims

1. A plant operation support device that supports plant operations, a prediction processing unit that predicts a change in a process value of the process by using a process model corresponding to the process of the plant; the prediction processing unit, when a manual intervention is performed by an operator of the plant after the process value is predicted to exceed a predetermined threshold, predicts a change in the process value taking into account the performed manual intervention, and also assumes a manual intervention different from the performed manual intervention and predicts a change in the process value taking into account the assumed manual intervention; a display processing unit that displays a first trend graph of the process value taking into account the performed manual intervention in a superimposed manner together with a second trend graph of the process value taking into account the assumed manual intervention.

2. 2. The plant operation support device according to claim 1, wherein the display processing unit displays the content of the performed manual intervention and the content of the assumed manual intervention in a comparative manner on a screen different from a trend screen on which the first trend graph and the second trend graph are superimposed.

3. the prediction processing unit predicts a start point at which the process value exceeds the threshold, and then predicts an end point at which the process value does not exceed the threshold due to the manual intervention; 3. The plant operation supporting device according to claim 1, wherein the display processing unit displays the first trend graph and the second trend graph in a section from the start point to the end point in an overlapping manner.

Images