JP3951374B2 - Plant interface agent - Google Patents

Description

[0001]
BACKGROUND OF THE INVENTION
The present invention relates to an interface agent for a plant, and more particularly to a technique for monitoring the operation state of a plant using an artificial intelligence (AI) technique.
[0002]
[Prior art]
As a general form, a plant includes process equipment that performs a certain process, control equipment that controls the process equipment, and operation equipment that allows an operator to give the process equipment some operation. For such a plant, in the operation of a conventional plant, the operator grasps the operation state of the plant by reading the measurement data (measured values measured for the process equipment) displayed on the display means of the operation equipment. Then, an instruction is given to the operation of the plant by performing operation input to the operation equipment as necessary.
[0003]
[Problems to be solved by the invention]
By the way, in the operation of a conventional plant, it is an operator's unilateral responsibility to grasp the operation state of the plant. That is, the operator grasps the operating state of the plant based on a plurality of measurement data displayed by operating the operating device. In order to perform an appropriate operation input, considerable skill is required. In particular, in an emergency such as when a failure occurs in a part of the plant, the operator needs to select and check a large number of measurement data more accurately than usual, so the load is extremely large. .
[0004]
It should be noted that the operator's work during normal times is centered on monitoring of the measurement data, and it is difficult to maintain concentration. Therefore, it is difficult for the operator to quickly and accurately grasp the change in the operation state of the plant.
[0005]
The present invention has been made in view of the above-described problems, and has the following objects.
(1) To provide a plant interface agent capable of assisting an operator to grasp the operation state of a plant and to input an operation to the plant.
(2) To provide a plant interface agent capable of realizing stable plant operation without depending on the skill of the operator.
(3) To provide a plant interface agent capable of quickly and accurately grasping changes in the operation state of a plant.
[0006]
[Means for Solving the Problems]
In order to achieve the above-mentioned object, in the present invention, as a first means, a message is provided between the plant and the operator and assists the operation of the plant based on the plant data indicating the operation state of the plant. In the plant interface agent that outputs to the plant, the plant data related to the plant operation is subjected to predetermined processing and sequentially stored in the case memory as time series data, and is similar to the current time series data indicating the current plant operation state A means comprising an inference engine for comparing and collating past time series data based on a case-based reasoning method and generating a message based on the result of the comparison and matching is employed.
As a second means, in the above means, the inference engine maps the absolute value of each plant data sampled at fixed time intervals into a metric space, and sequentially stores mapping information obtained by the mapping as time series data. , Calculating the distance between past time-series data and current time-series data for each plant data based on the mapping information, detecting past time-series data having a small sum of the distances as similar operation examples, A means is adopted that is configured to generate a message to the operator based on a similar driving case.
As a third means, in the second means, the inference engine calculates a change rate of each plant data, maps the change rate to a metric space, and changes based on time data attached to each plant data. A means for associating a rate with the absolute value, and storing the mapping information relating to the mapping of the absolute value together with the mapping information relating to the mapping of the change rate and the associating information by the association as time series data Is adopted.
As a fourth means, in the second or third means, when the inference engine detects at least one absolute value or change rate that is a distance larger than a certain value, the past time series data is similar. A method is adopted in which it is configured so as to be excluded from the driving case candidates and omit the distance calculation of other absolute values or change rates related to the past time series data.
As a fifth means, in the first to fourth means, means having command output means for outputting the operation information input from the operator as a command to the plant is adopted.
As the sixth means, in each of the first to fifth means, as plant data, among the measurement data measured for various process devices constituting the plant and the operation operation data operated by the operator on the plant, A means of adopting either one or both is employed.
[0007]
DETAILED DESCRIPTION OF THE INVENTION
Hereinafter, an embodiment of a plant interface agent according to the present invention will be described.
[0008]
FIG. 1 is a block diagram showing a functional configuration of the plant interface agent of the present embodiment. In this figure, reference numeral 1 denotes an interface agent for the plant, which is interposed between the plant and the operator and supports information exchange between the plant and the operator.
[0009]
As a general form, a plant includes process equipment that performs a certain process, control equipment that controls the process equipment, and operation equipment that allows an operator to give the process equipment some operation. The operator grasps the operating state of the plant by reading the measurement data (measured values measured for various process devices) displayed on the display means of the operating device, and operates the plant by inputting operation to the operating device. Give instructions to.
[0010]
The plant interface agent 1 is provided, for example, in parallel with the operation device, and assists in grasping the operation state of the plant based on the measurement data and operating input to the plant. As will be described below, the plant interface agent 1 applies the case-based reasoning, which is a kind of artificial intelligence (AI) technique, to make the operation state of the plant as a past operation case (past time series data). It accumulates sequentially in time series, and determines the current operating state based on the past time series data.
[0011]
The plant interface agent 1 includes a user interface 1a, a language conversion system 1b, an inference engine 1c, a matching rule 1d, a case memory 1e, a command execution module 1f, a data collection module 1g, and the like.
[0012]
The user interface 1a is, for example, a display, a speaker, a keyboard, a voice input device, and the like, and outputs operation information input from an operator by operating the keyboard or by voice input to the language conversion system 1b as operation data. The language data input from the language conversion system 1b is converted into an audio signal and converted to a speaker, or converted into a video signal and output to a display.
[0013]
The language conversion system 1b converts the operation data input from the user interface 1a into a computer language that can be processed by the inference engine and outputs the computer language to the inference engine 1c. The language conversion system 1b understands the computer language input from the inference engine 1c. This is converted into linguistic data that can be output to the user interface 1a.
[0014]
The inference engine 1c generates current time series data indicating the current operating state by processing the plant data input from the data collection module 1g using a case-based reasoning method, and is similar to the current time series data. The past time series data is searched from the case memory 1e based on the matching rule 1d, and the computer language of the operation data input from the language conversion system 1b is output to the command execution module 1f as it is. The inference engine 1c sequentially stores the current time series data generated based on the plant data in the case memory 1e as past time series data.
[0015]
The case memory 1e stores current time series data sequentially input from the inference engine 1c as past time series data. The collation rule 1d shows a collation procedure between the current time series data and the past time series data based on the case-based reasoning technique. The processing of the inference engine 1c based on the matching rule 1d will be described in detail below.
[0016]
The command execution module 1f converts the computer language of the operation data input from the language conversion system 1b via the inference engine 1c into a command (instruction) that can be understood by the plant control device and outputs the command to the plant. The data collection module 1g collects the command and the command input to the plant via the operation device and the measurement data as plant data, and outputs the collected data to the inference engine 1c.
[0017]
Next, the operation of the plant interface agent 1 configured as described above will be described with reference to the flowchart shown in FIG.
[0018]
First, in step S1, the data collection module 1g samples the operation data and the measurement data at regular time intervals, whereby plant data is input to the plant interface agent 1. In this case, the plant data is formed from measurement data over a plurality of physical quantities and a plurality of operation information related to various process devices.
[0019]
The plant data input in this way is input to the inference engine 1c and undergoes the following processing. That is, the absolute value of each plant data is mapped to the metric space (step S2). For example, as shown in FIG. 3, the mapping of the absolute value to the metric space includes plant data related to Steam Flow (steam flow rate) in each tree-structured link (Steam Flow.0 to Steam Flow.6). (Mapping process) to be assigned to Here, each link is obtained by dividing the maximum fluctuation range of the steam flow rate into several (in this case, six) small fluctuation ranges.
[0020]
For example, in FIG. 3, when 0.58 is newly sampled, the plant data is Steam Flow. Assigned to 6 links. That is, the absolute value of the plant data sampled sequentially with the passage of time is mapped to one of the links according to the value, and is mapped to the metric space. Such mapping of absolute values to the metric space is performed for plant data of all physical quantities.
[0021]
When the mapping of the absolute value of the plant data to the metric space is completed in this way, a label indicating the characteristics of the plant data is assigned to each plant data (step S3). For example, in the case of plant data related to the above steam flow, labels such as “large steam flow”, “slight steam flow”, “normal steam flow”,... To grant.
[0022]
On the other hand, in step S4, the rate of change is calculated by comparing the previously sampled plant data with the current plant data. In a plant, not only the absolute value of plant data but also its tendency to change is often an important indicator. Therefore, the rate of change is calculated by normalizing the plant data with the maximum value and taking the difference between the normalized plant data (normalized plant data) and the normalized plant data of the previously sampled plant data. The change rate is calculated for all physical quantities.
[0023]
When the change rate of the plant data is calculated in this way, the change rate is mapped to the metric space (step S5). The mapping of the absolute value to the metric space is performed in the same manner as the absolute value described above, and the rate of change sequentially calculated with the passage of time is set for each of the tree structures predetermined for the rate of change. It is a process assigned to a link.
[0024]
Then, when mapping of the change rate of the plant data to the metric space is completed, labeling is performed (step S6). For example, in the case of the above steam flow, depending on the rate of change of the steam flow, labels such as “high rate of change of steam flow”, “slightly high rate of change of steam flow”, “normal rate of change of steam flow”, etc. Is given for the rate of change of each physical quantity.
[0025]
As described above, the absolute value and the change rate of the plant data are independently mapped to the metric space and labeled. In step S7, the mapping information of the absolute value to the metric space and the change rate to the metric space are converted. Mapping information correlation processing is performed (step S7). In this correlation process, the rate of change and the absolute value are related based on time data associated with sequentially sampled plant data.
[0026]
When the labeling of the absolute value and the change rate of the plant data and the correlation process are completed in this way, the series of processing results, that is, the mapping information and the labeling information of the absolute value and the change rate of the plant data to the metric space. The related information is the current time series data described above, and is stored in the case memory 1e in time series (step S8).
[0027]
That is, current time series data is generated based on plant data sequentially sampled with the passage of time from the start of plant operation by the data collection module 1g, and stored in the case memory 1e as past time series data indicating past operation cases. Accumulated and stored in time series.
[0028]
Then, the inference engine 1c matches one or a plurality of past time series data over a certain past time period already stored in the case memory 1e against the current time series data of the newly sampled plant data. By comparing and collating according to the above, a past driving case similar to the current driving case is searched as a similar driving case (step S9).
[0029]
Specifically, the mutual distance between the data is calculated for all combinations of one or more past time-series data and current time-series data over a certain past time span. That is, the distance between the past time-series data and the current time-series data is calculated for all absolute values and change rates mapped to each link by the mapping process to the metric space (steps S2 and S5). Then, the past time series data having the smallest sum of the distances is detected as a similar operation case most similar to the current time series data.
[0030]
For example, in FIG. 3, the absolute value of the plant data belonging to the same link has the smallest distance. And the distance which is in the relation which reaches | attains following more links becomes large. Steam Flow. The absolute value of 0.34 belonging to link 4 is Steam Flow. The absolute value 0.44 belonging to the link 5 is Steam Flow. 2 links can be followed.
[0031]
In contrast, Steam Flow. The absolute value of 0.34 belonging to link 4 is Steam Flow. The absolute value of 0.68 belonging to the link of 3 is Steam Flow. It can be traced via a zero link. Therefore, the absolute value 0.44 has a smaller distance than the absolute value 0.68 with respect to the absolute value 0.34 because the number of links that need to be traced is small.
[0032]
Such distance calculation of each plant data is performed for all combinations of one or more past time-series data and current time-series data over a certain past time width. However, in order to shorten the processing time, if at least one absolute value or change rate that is a distance greater than a certain value is detected, the past time series data is excluded from similar operation case candidates, It is conceivable to omit distance calculation of other absolute values or change rates related to past time series data.
[0033]
When the similar operation case is detected in this way, the inference engine 1c infers the current operation state of the plant based on the past time series data related to the similar operation case, and generates the inference result message data. For example, when the similar operation case is stored in the case memory 1e when the specific device constituting the plant is in an abnormal state in the past, the current time-series data indicates a plant abnormality. The inference engine 1c generates message data that informs the operator of this fact.
[0034]
This message data is, for example, a computer language that can be output by the inference engine, and is output to the language conversion system 1b. The language conversion system 1b converts the message data into a language that can be understood by humans, for example, data such as Japanese or English, and outputs the data to the user interface 1a. As a result, a message notifying the abnormality of the specific device is output from the user interface 1a to the operator in the form of voice or video (step S10).
[0035]
On the other hand, when an operation input is made from the operator to the plant interface agent 1 in response to a message notifying the abnormality of the specific device, the operation input is input to the language conversion system 1b via the user interface 1a. And converted to computer language. The computer language is input to the command execution module 1f via the inference engine 1c and converted into a command that can be handled by the plant control device. This command is output to the plant as an operation instruction.
[0036]
For example, this command may be a command that instructs the specific device to stop operating or a command that changes the operating state of the specific device. As described above, the command output from the plant interface agent 1 to the plant and the command input from the operation device originally provided in the plant are used as one of the plant data in the data collection module 1g. Sampled.
[0037]
In the above embodiment, measurement data and operation data are taken up as plant data. However, the present invention is not limited to this, and either one may be handled. The time width set for past time series data for comparison with the current time series data and which physical quantity of plant data to compare with respect to past time series data and current time series data are monitored. It is necessary to set as appropriate based on the characteristics of the plant.
[0038]
【The invention's effect】
As described above, the plant interface agent according to the present invention has the following effects.
(1) A plant interface agent that is interposed between a plant and an operator and outputs a message for supporting the operation of the plant to the operator based on plant data indicating the operation state of the plant. The data is processed and stored in the case memory as time-series data in sequence, and past time-series data similar to the current time-series data showing the current plant operating state is compared with the case-based reasoning method. Since the reasoning engine which compares and collates using and produces | generates a message based on this comparison and collation result is provided, the grasping | ascertainment of the operation state of the plant by an operator and the operation input to a plant can be supported effectively.
(2) Since a message for supporting the operation of the plant for the operator is output based on the plant data, stable plant operation can be realized without depending on the skill of the operator.
(3) Since a message for supporting the operation of the plant is automatically generated based on the time series data accumulated sequentially, the message reflects the operation state of the plant quickly and accurately. Therefore, the operator can grasp the change in the operation state of the plant quickly and accurately.
(4) By using the operation data as the plant data to be input to the case-based reasoning, it is possible to monitor the operation of the operator for the plant and output a message for it.
(5) By applying case-based reasoning not only to the absolute value of plant data but also to the rate of change thereof, it is possible to more accurately monitor the operating state of the plant and output a message.
[Brief description of the drawings]
FIG. 1 is a block diagram showing a functional configuration of an embodiment of a plant interface agent according to the present invention.
FIG. 2 is a flowchart showing the operation of an embodiment of a plant interface agent according to the present invention.
FIG. 3 is an explanatory diagram illustrating mapping of current time-series data to a metric space in an embodiment of a plant interface agent according to the present invention.
[Explanation of symbols]
DESCRIPTION OF SYMBOLS 1 ... Plant interface agent 1a ... User interface 1b ... Language conversion system 1c ... Inference engine 1d ... Collation rule 1e ... Case memory 1f ... Command execution module (command output means)
1g …… Data collection module

Claims (7)

An interface agent for a plant that is interposed between a plant and an operator and outputs a message for supporting the operation of the plant to the operator based on plant data indicating the operation state of the plant,
Predetermined processing is performed on plant data related to plant operation, and time-series data is sequentially stored in the case memory (1e), and past time-series data similar to the current time-series data indicating the current operation state of the plant is used as an example. An inference engine (1c) that performs comparison and collation based on a base inference method and generates a message based on the comparison and collation result;
The inference engine maps the absolute value of each plant data sampled at fixed time intervals into a metric space by mapping it to a predetermined tree structure, and sequentially maps information obtained by the mapping as time series data. Storing the distance between past time-series data and current time-series data for each plant data based on the mapping information, and detecting past time-series data with a small sum of the distances as similar operation cases. Generating a message to the operator based on the similar driving case;
An interface agent plant for plants. The plant interface agent according to claim 1,
The inference engine calculates a change rate of each plant data, maps the change rate to a metric space, relates the change rate and the absolute value based on time data attached to each plant data, and A plant interface agent characterized in that mapping information relating to mapping of values and mapping information relating to mapping of change rate and association information by association are stored as time series data. 3. The plant interface agent according to claim 1, wherein the inference engine detects the past time-series data of a similar operation case when at least one absolute value or change rate having a distance larger than a certain value is detected. An interface agent for a plant, which is excluded from candidates and omits distance calculation of other absolute values or change rates related to the past time series data. The plant interface agent according to any one of claims 1 to 3, further comprising command output means (1f) for outputting operation information input from an operator to the plant as a command. The plant interface agent according to any one of claims 1 to 4, wherein the plant data is measurement data measured for various process devices constituting the plant. The plant interface agent according to any one of claims 1 to 4, wherein the plant data is operation data operated by an operator on the plant. The plant interface agent according to any one of claims 1 to 4, wherein the plant data is measurement data measured for various process devices constituting the plant and operation operation data operated by an operator on the plant. Feature plant interface agent.

Images