JP3443787B2 - Autonomous decentralized plant control man-machine device - Google Patents

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a system for operating a controller for controlling plant equipment via a plurality of man-machine devices, without requiring a dedicated hardware server device. A plant control man-machine that enables a machine device to independently output control data to a controller (control device) and prevents the control data from competing and outputting to the same control target (controller). Regarding the device.

[0002]

2. Description of the Related Art Conventionally, in order to operate a plant device, a controller (control device) for controlling each device and a man-machine device for giving an instruction to the controller are connected by a network connection device, and the man-machine device is connected to the control device. A system for performing an output command operation is already widely known and put into practical use.

[0003]

However, in the above-mentioned conventional plant control system, a dedicated hardware server is installed, and this and the man-machine device are connected to the network by using a plurality of network connection devices. In some cases, a dedicated hardware server is not installed, and a plurality of man-machine devices are connected to the network using a network connection device.

In the former case, if the server device fails, the plant control operation from the man-machine device cannot be performed. Therefore, two servers must be installed and duplicated, and when one server fails, Until the backup of the standby system is completed, there is a drawback that the plant control operation cannot be performed from the man-machine device.

In the latter case, when a plurality of man-machine devices perform control operations on the operating end of the same controlled object, operation conflicts occur, and commands from other man-machine devices arrive later. There is a drawback in that the operation of the control target operation end may be inconsistent with the intended operation in some cases.

[0006]

Therefore, according to the present invention,
At least one control device for controlling the input device;
Multiple man-machine devices that output control data to the control device
Between the control device and the plurality of man-machine devices
Includes a network transmission line for transmitting and receiving data
Independent self-contained plant control man-machine equipment composed of
In addition, the man-machine device is
Data exchanged with the
Mass that indicates whether the machine is a master or slave
Common memory where the flag is stored and its own man-machine
Write the control data output from the device to its own common memory
Write to the common memory of other man-machine devices
Each has a network control device
Refer to the master flag of another man-machine device
If another master already exists, the master
If the flag is set to slave state and no other master exists,
Communication with other man-machine equipment, the priority of mastering
If the self should become the master according to
The master flag is set to the master state, and the control device
Refer to the master flag of the common memory of the machine
In the common memory of the man-machine device in the master state,
The stored control data is read and the plant machine
It is characterized by controlling the vessel .

As described above, each man-machine device has a control device.
Set up a common memory to store data to and from
Control data output from its own man-machine device.
Write to your own common memory and
Since it was written in the common memory of
Control data with the same contents is stored in the shared memory of the machine.
Will be paid. In addition, a common mechanism for each man-machine device
Store the master flag in memory and
Refer to the master flag of another man-machine device,
If there is a master in the
If there is no other master, the other
Communicates with the machine device and follows the priority of mastering
If you want to become the master,
Turn it on.

This eliminates the need for a dedicated server,
Each man-machine device is independent of the control device
It becomes operable. In particular, the controller should refer to the master flag.
The man-machine device in the master state,
Control data stored in the common memory of the machine
Since it can be read out and used to control plant equipment, the same
Device with multiple control data for each control target
It is possible to prevent the occurrence of operation conflict due to the output from
It

The common memory is a man-machine device of its own.
Memory is updated when the device is operating normally.
It has a rear and man-machine device is
Based on the contents of the storage area, the other man-machine device
The man-machine device is not normal.
And the master flag is in the master state,
With other man-machine devices except the other man-machine device
Communicate and become the master according to the priority of mastering.
If it should, set its own master flag to the master state.
You can In other words, the failure of the man-machine device
Automatically detected and operated as a master until then
Man-machine device that is alive even if the man-machine device fails
The master is automatically switched to. Moreover, as mentioned above
As a result, which man-machine device became the master
However, the same control data is stored in each shared memory.
Affected by the failure of the man-machine device
Instead, the operation to the control device can be continued.

[0010]

BEST MODE FOR CARRYING OUT THE INVENTION Embodiments of the present invention will be described in detail below with reference to the accompanying drawings. FIG. 1 shows the overall configuration of a self-sustained distributed power plant control man-machine apparatus according to an embodiment in which the present invention is applied to a power plant.

The self-sustained distributed power plant control man-machine device shown in FIG. 1 includes a plurality of control devices (only one unit is shown in FIG. 1) 3 including a controller for controlling plant equipment, and this control device. 3 connected to the network 4 via the network control device 2, and a plurality of man-machine devices 1 connected to the network 4 via the network control device 2 and outputting control data to the control device 3.
And a CRT display 1A and a keyboard (K / B) 1B connected to each man-machine device 1.

The man-machine device 1 and the control device 3 are respectively connected to the network 4 via the network control device 2, and the operation command to the control device 3 is sent to the CRT display 1A connected to the man-machine device 1. K / B1B
Is output to the network control device 2 via the man-machine device 1. The operation command output to the network control device 2 is sent to the network 4,
From the network 4, the network control device 2 of the control device 3 takes in the control device 3.

Next, a method of realizing a common memory between network control devices will be described with reference to FIG. In FIG. 2, by periodically matching the internal memories 21 of the network control devices 2, it is possible to secure a common memory 22 that is a common memory area among a plurality of network control devices 2. By reading and writing this common memory 22,
Data can be exchanged between the plurality of network control devices 2.

It will be described with reference to FIG. 3 that each man-machine device can independently exchange data with the control device by dividing the unit for reading and writing the common memory 22 for each device. As shown in FIG. 3, when the man-machine device 11, the man-machine device 12, and the control device 3 are connected to the network 4, the common memory 22 is divided into three, and the man-machine device 11 and the man-machine device 12 are divided into three parts. , And the write-only area of the control device 3,
The man-machine device 11 and the man-machine device 12 can output data to the control device independently of each other. However, even though the output data from the man-machine device 11 and the man-machine device 12 are independent, if the man-machine device 11 and the man-machine device 12 output data independently to the same control target, Therefore, it is not possible to prevent contention of the output of the same control target data.

In order to prevent this, master device management of man-machine devices described below is performed, and when two or more man-machine devices are connected to the network, one man-machine device becomes the control device. By controlling the data output, the data output conflict is prevented.

In the following description, the man-machine device will be referred to as C
It is abbreviated as PU, and the target CPU is the main CPU, the partner CPU is the partner CPU, and C is used when the CPU is not particularly limited.
A state in which PUi and CPUi are master devices is called a master, and a state in which they are not master devices is called a slave.

With reference to FIG. 4, a method of determining whether the CPUi is alive or dead will be described. To determine whether the CPUi is alive or dead, counters 2211 and 2221 that are constantly updated when the CPUi is operating normally (live) are provided in the common memory of the CPUi. C
The life / death counter 2211 of the common memory previous read information 221 of PUi and the life / death counter 2221 of the common memory current read information 222 of CPUi are compared. If they match, the dead state of CPUi is set. The mastering priority is set to CPUi> CPUi + 1,
The CPUi to be the master is determined based on the life-and-death state of the CPUi and the priority order of the mastering.

FIG. 5 shows a procedure for deciding a CPU to be a master when there are two CPUs. When the CPU is started up, it is first determined whether or not a master already exists (P50). If a master already exists, the process proceeds to P56, the fact that it is in the slave state is recorded in the master flag area in the common memory 222 in FIG. 4, and this CPU is made the slave. If the master does not yet exist, the process proceeds to P51 to report the mastering of this CPU to CPUi, and P52
Wait for a response in the delay process of. In the process of P53, the state (live or dead) is determined by the presence or absence of the response of the CPUi, and if the CPUi is in the dead state, this CPU is started by the master CP
When it is determined to be U, the master state is set in process P55, and the fact that it is in the master state is recorded in the master flag area in the common memory 222 in FIG.

If the CPUi is in the raw state, the response reported to all the CPUs is judged at P54. If this response is a mastering affirmative response (acknowledgment response to the fact that this CPU becomes a master), it means that another CPU has given up because the mastering priority is low. At this time, this CPU becomes the master in the process of P55.

Next, the response to the masterization report will be described with reference to FIG. By the process of P51 in FIG. 5, all C
When a masterization report is made to the PU and a response is requested, the CPU that received the report is notified of P61 to P64 in FIG.
Follow the procedure in. First, in P61, the CPU and the CPUi that issued the mastering report are compared based on the mastering priority order to determine the mastering priority order. If the priority order of this CPU is low, the process proceeds to P63 to make an affirmative report to CPUi, and then makes itself a slave in the process of P64. On the other hand, if the priority of this CPU is high, P
By proceeding to 62 and making a negative report to CPUi, the first startup CPU becomes the master CPU,
When both CPUi and CPUi are booted at the same time, the mastering concession based on the mastering priority is guaranteed.

Further, in FIG. 1, the man-machine device 1 changes the state of another man-machine device 1 to the common memory 221 of FIG.
By managing life and death using the life and death counters 2211 and 2221 shown in and 222, the master of the man-machine device 1 can be switched even when the man-machine device fails. FIG. 7 shows an example of the procedure of life-and-death management executed by all man-machine devices at a predetermined time interval during the operation of the device. First, in the process of P71, it is determined whether or not this CPU is in the raw state and the master. If it is in the raw state and the master, the master state is held. In other states, it is determined in the process of P72 whether or not there is another raw master CPUi. If there is another raw master CPU i, that state is maintained. Raw Master C
If there is no other PUi, the process proceeds to P51 in FIG. 5 to follow the procedure to switch the main CPU to the master state.

Next, with reference to FIG. 3, description will be given of a case where the man-machine device 11 which is the master and the man-machine device 12 which is the slave output control data independently of each other.

The control data output from the man-machine device 11 is written in the memory 21A of the network control device 2A, and at the same time, is stored in the memory 21B of the network control device 2B of the man-machine device 12 by the message communication function of the network control device 2A. Written. On the contrary, the control data output from the man-machine device 12 is
It is written in the memory 21B of the network control device 2B, and at the same time, written in the memory 21A of the network control device 2A of the man-machine apparatus 11 by the message communication function of the network control device 2B. Therefore, the memories 21A and 21B always store the same content, and the man-machine device 11 and the man-machine device 12 share one memory.

For one control operation of the same device,
When the control data is output from the man-machine device 11 and the man-machine device 12, the control data output first is written in both memories, and is output later until the operation based on the control data ends. The control data cannot be written in the memory. Therefore, two man-machine devices do not compete for certain operations of the same device.

On the other hand, the network control device 2C of the control device 3 has the memory 2 whenever the writing to the memory 21A or the memory 21B is performed or at a predetermined time interval.
Read the master flag area of 1A and memory 21B,
Which one is in the master state is checked, the control data written in the memory in the master state (memory 21A in this case) is read and sent to the control device 3. Based on the control data sent from the control device 3,
Controls each plant equipment. The control device 3 further sends the control result (device state) and the process amount to the network control device 2C, and the network control device 2C writes the data in the memory 21C. The man-machine devices 11 and 12 read the contents of the memory 21C periodically or as needed, and perform necessary processing.

As described above, the memory 21A and the memory 21
Since the stored contents of B are always the same, even if the man-machine device 11 which is the master goes down, the master is simply switched to the man-machine device 12, and the control is continued using the memory 21B. You can

[0027]

According to the self-sustaining distributed plant control man-machine device of the present invention, 1) the man-machine device can output operation commands to the control device independently of each other, so that a dedicated server is not required. 2) It becomes possible to construct a plant control man-machine device in which failure of other man-machine devices does not affect the operation of the control device. 3) Master even for operation requests to the control device from multiple man-machine devices. Since the man-machine device of the output source can be unified by performing management, it is possible to construct the plant control man-machine device with less erroneous operation due to competition.

[Brief description of drawings]

FIG. 1 is an overall configuration diagram of an autonomous distributed power plant control man-machine apparatus according to an embodiment of the present invention.

FIG. 2 is a diagram illustrating a concept of a common memory of the network control device in the embodiment of the present invention shown in FIG.

FIG. 3 is a diagram showing an example of memory division of the common memory shown in FIG. 2 for each device.

FIG. 4 is a diagram showing a concept of a life-and-death management method of a man-machine device using a common memory.

FIG. 5 is a procedure diagram showing an operation of a mastering determination function in the present invention.

6 is a procedure diagram showing a procedure of responding to a masterization report of the procedure of the masterization determination function shown in FIG.

FIG. 7 is a procedure diagram showing master switching to which the mastering determination function shown in FIG. 5 is applied when the man-machine device fails.

[Explanation of symbols]

1 Man-machine device 2, 2A, 2B, 2C Network control device 3 Control device 4 Network transmission line 11, 12 Man-machine device 21, 21A, 21B, 21C Network control device internal memory 22 Common memory 221 CPUi common memory Previous read information 2211 CPUi common memory previous read information live / dead counter 222 CPUi common memory current read information 2221 CPUi common memory current read information live / dead counter

─────────────────────────────────────────────────── ─── Continuation of the front page (56) Reference JP-A-7-152426 (JP, A) JP-A-9-160636 (JP, A) JP-A-2-114294 (JP, A) JP-A-9- 116591 (JP, A) JP-A-9-34540 (JP, A) (58) Fields investigated (Int.Cl. ⁷ , DB name) G05B 15/02 G05B 9/03 G06F 15/16 H04L 12/40

Claims (2)

(57) [Claims] 1.At least one controlling plant equipment
Of the control device and a device for outputting control data to the control device.
Number of man-machine devices, the control device and the plurality of man-machine devices.
A network that transmits data to and from machine devices.
Transmission lineIndependent distributed plant system
In your man-machine equipment,The man-machine device includes the control device and the man-machine.
Data exchanged with the device and its own mamma
Master flag indicating whether the device is a master or a slave
From a common memory that stores
When the output control data is written in its own common memory,
In addition, the network that writes to the common memory of other man-machine equipment.
Each has a network control device and
Refer to the master flag of the man-machine device of
If a master exists in the
If there is no other master in the rave state, another
Computer and communicates with the machine device according to the priority of mastering.
If you want to be the master, set your own master flag
To master state, The control unit controls the common memory of the man-machine units.
Refer to the master flag of
Control data stored in the common memory of the
It is characterized by reading out and controlling the plant equipment.
Do Independent distributed plant control man-machine equipment. 2.The common memory is its own man-machine device.
Memory is updated when the device is operating normally.
Have a rear, The man-machine device is the memory of another man-machine device.
The other man-machine device is normal based on the contents of the area.
Whether the man-machine device is not normal, and
When the master flag is in the master state, the other
Communicate with other man-machine devices except man-machine device
If you should become the master according to the priority of mastering
If the master flag is set to the master state,
The method according to claim 1, characterized in that Independent distributed plant control machine
Machine equipment.