JPS58140803A - Process controller - Google Patents

Abstract

PURPOSE:To obtain a process controller of double structure which has a simple and assured operation, by combining two control arithmetic parts, a memory containing an error correcting function and a switch control part. CONSTITUTION:Control arithmetic parts 1A and 1B, a switch control part 7, an input/output part 4 and a controlled system 5 are formed centering on a memory 8 containing an ECC. The memory 8 stores all kinds of information necessary for control. As error of the memory 8, if arises while the part 1A is under a control operation, produces no problem to continue the control as long as it is within a correctable range. In case the part 1A has a fault, the part B is connected to the memory 8 by the control part 7. Thus the part 1B continues the control operation on the basis of the contents of the memory 8. As a result, a double-structure process controller having a simple and assured operation can be obtained.

Description

DETAILED DESCRIPTION OF THE INVENTION The present invention relates to a process control device, and in particular to the configuration of a process control device including sequence control in which control history is a problem. When high performance is required, it is common to configure multiplexed (in most cases duplex) configurations. 1st
Figures and! 211t;j is a diagram showing two examples of its configuration.

In Figure 1, the A and Br1i control systems are connected using a common memory 3t, and if, for example, the A control system fails, information in the middle of the control (for example, how far the calculation has been performed) will be stored in the common memory 3. The system is configured as software, and the B control system executes attack control based on this information. Here, the operation of each part will be described in the case where the AIItlJm system is operating and the control system is shifted to the B control system.

The control system consists of a control calculation section IA and a dedicated memo IJ2A, and the control calculation section IA is contained in the dedicated memo IJ2A. A 1tilJ control operation is performed on the controlled object 5 through the input/output unit 4 based on the I/N information (so-called program).At this time, a new calculation result is obtained for information that requires a history, such as information in the middle of calculation. It is stored in the common memory 3 every time.In such a state, if a failure occurs in the XO control section IA, the result is transmitted to the switching control section 7 as a diagnostic IW report 11A.
Based on the information, the 11jtiL1 unit 7 performs an operation to switch the control by the 1fflJ control calculation unit IA to the control by the control calculation unit IB. Control calculation mlB is executed in dedicated memory 2BCl
The same contents as the memory IA are stored in the control target 3 & (in contrast, the human output unit 4
Control m @ production through.

In this way, it is possible to continue the predetermined control operation even if trouble occurs in any of the control calculation units.

The configuration shown in FIG. 2 is almost the same, but unlike the case shown in FIG. 1, the common memory 3 is replaced by a communication device 6. In the case of FIG. 2, the history information to be stored in the common memo I73 is immediately sent to the dedicated memory of one control system via the communication line 6, and the common memory is divided into dedicated memories IA and IB, respectively. Similarly to FIG. 1, if a failure occurs in either control calculation section IA or IB, control can be continued with one control system.

However, in such a configuration, the memory is divided into two or three pieces in terms of hardware, which makes the configuration complicated.
It is necessary to store both the data and the IB. This reliability means that the contents of both memories must always be the same when storing or flying a program, or setting or changing parameters, which has the disadvantage of complicating handling.

In the case of Fig. 1, two types of memory are required in terms of hardware, while in the case of Fig. 2, the A and B control systems need to be constantly overconfident, and the system configuration is complicated in both cases. It has the disadvantage of becoming

An object of the present invention is to provide a redundant process control device that is simple in configuration and reliable in operation.

Conventionally, one of the ideas for backing up a control device is to multiplex (in many cases, duplicate) a failed part and continue control operations. In other words, there is no need to back up parts that do not fail.Therefore, the present invention uses a memory with air correction (hereinafter referred to as memory with ECC) in the memory section, which has had many problems in the past. Based on the idea of not considering memory backup in case of a major failure, we took the idea of duplicating only the control calculation section, and created an effective duplex process @ device with a simple configuration. It is something. In other words, the process control device is constructed based on the idea that parts that can be repaired by themselves will be self-repaired, and only parts that cannot be repaired will be duplicated.

The present invention will be explained below based on examples with reference to FIGS. 3 and 4.

FIG. 3 is a block diagram showing an embodiment of the present invention, and FIG. 4 shows the main parts extracted to explain bus switching.

In these figures, the system is composed of control calculation sections IA and IB, a switching control section 7, an input/output section 4, and a controlled object 5, with the memory 8 with ECC as the center. The memory with ECC 8 stores all information necessary for control (programs, parameters, intermediate results of calculations, etc.). Now control calculation unit IA
Suppose that is performing control calculations. In this case, even if an error occurs in the ECC memory 8, it is within the correctable range (for example, 1
If the bit error is repaired or a 1-bit error occurs when a 2-bit error is detected, for example, the 0-time yII will only generate an alarm and will not cause any interruption in the continuation of control.
This is a case where a failure occurs in the IFi calculation control unit IA. In this case, based on the diagnostic report 11A, the switching control unit 7 is disconnected from the control calculation unit IAt memory bus 9 by the switching signal 14A, and connected to the control calculation unit IBt memory bus 9 by the switching signal 14B.

This allows the control calculation section IB to continue the control calculation according to the contents of the ECC memory 8 that stores all control information.

FIG. 4 shows an example of the connection method between the control calculation units IA and IB and the memory with ECC 8, in which only the output data 12A and 12B of the control calculation units IA and IB are connected to the party switching signals 14A and 14.
Output game by B) 13A.

13B.

In this way, according to this embodiment, one memory with ECC 8
It is possible to switch between the two control calculation units only by using , and compared with the conventional method shown in FIGS. That is, (1) there is no need to think about backing up the memory, and only the control calculation section needs to be backed up, which simplifies the hardware configuration.

(2) When switching control systems or changing programs, it is only necessary to consider only one memory, which eliminates the need to always consider two or more memories as in the conventional method, and the software becomes simpler.

As is clear from the above description, according to the present invention, the backup of the process control device can be limited to only the control calculation section, so that the following effects can be achieved. That is, since only one type of memory is used, the system configuration is simplified and handling from both hardware and software aspects is simplified. In particular, since all information including control information is stored in one memory, the transfer of control calculations is extremely clear.

In addition, for standby control calculations, it is sufficient to always start the depth from the initial state.In addition, although a microphone processor is often used in the control calculation section, it is only necessary to switch the output signal. Very effective results can be obtained, such as the ability to use a switching method that does not depend on processor architecture.

[Brief explanation of the drawing]

1 and 2 are block diagrams of a conventional device, FIG. 3 is a block diagram showing an embodiment of the present invention, and FIG. 4 is a block diagram showing the main parts of the above embodiment. 1A, IB... Control calculation unit, 2A, 2B... Dedicated memory, 3... Shared memory, 7... Switching control unit, 8.
...FCC No. 1 No. 2 Fig. 3 Fig. 7 No. 4

Claims (1)

[Claims] 1. It consists of two control calculation units, one memory with an error correction function, and a switching control unit, and all information related to control is stored in the memory with an error correction function, and if a failure occurs in the control calculation unit or the memory. A process control device characterized by continuously performing four control calculations.