KR100416201B1 - Information processor and real-time distributed processing system - Google Patents

Abstract

In order to suppress the possibility of stopping the protection relay function and assign a plurality of tasks for realizing the protection relay function to the calculator, the present invention processes a plurality of tasks for realizing the protection relay function with a plurality of computers connected by a network. An information processing apparatus comprising: means for storing a combination of tasks to be processed by a specific operator among a plurality of tasks and a combination of tasks to be processed by the specific calculator to select the task of the selected combination; Means for executing with a calculator.

Description

Information processing device and real-time distributed processing system {INFORMATION PROCESSOR AND REAL-TIME DISTRIBUTED PROCESSING SYSTEM}

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a configuration of a distributed processing system having a plurality of computing devices connected to a network for executing a plurality of programs, and more particularly to a processing system for requesting real-time control represented by a protection control device of a power facility.

In a conventional general distributed processing system, various task assignment methods have been devised and actually used to improve processing capacity and reliability. For example, Japanese Patent Laid-Open No. 05-257908 and Japanese Patent Laid-Open No. Hei 07-282013 propose a method of assigning a task in accordance with the characteristics and states of each calculator. Task assignments and priorities are determined by referring to performance such as processing power and reliability of each of the operators, or states such as computational load. Tasks of different importance and urgency are handled by a calculator having appropriate reliability, so that highly reliable distributed processing is performed as a whole.

In addition to the above-described configuration, a mechanism of cooperative processing can be used as a correlation of tasks. For example, Japanese Patent Laid-Open No. 09-231183 provides a control mechanism for realizing time constraints on task cooperative processing such as simultaneous starting of tasks. However, in terms of reliability, task correlation is not solved.

On the other hand, as a real-time processing device that requires reliability, countermeasures by multiplexing have been conventionally taken. For example, in many control devices such as power control such as power generation, power transmission, and distribution, and traffic control such as aviation and railroad, the prevention of malfunction is strictly limited, but real time is required. Therefore, in such real-time processing, high reliability is realized by multiplexing control objects, multiplexing operations, multiplexing algorithms, and the like.

As a result, when the above-described object is provided as a highly reliable real-time distributed processing, a correlation is generated between the respective tasks to be distributed in terms of reliability. For example, when reliability is realized by providing a plurality of multiplexed tasks for the same purpose, it is preferable from the viewpoint of reliability that the task is not processed by the same computing device. Therefore, in the highly reliable real-time distributed processing, it is necessary to consider not only the correlation of the task in terms of the computation amount but also the correlation of the task in terms of reliability.

In the past, when a real-time processing device requiring reliability is to be installed in a distributed process, such relative constraints have to be satisfied by prior design. For example, Japanese Patent Laid-Open No. 57-3516 proposes a method of installing a protection control device of an electric power plant in a distributed processing system. However, adjustment of the amount of calculation and adjustment of the reliability are performed by human hands at the time of design. Therefore, it is necessary to design an optimum arrangement for each treatment device configuration. In addition, since only a predetermined task arrangement is used, the effect is lost when the processing apparatus configuration is changed.

As described in "Prior Art", the conventional general distributed processing system cannot be applied to a control device requiring reliability.

Previous distributed processing techniques prescribe the reliability and urgency of each operator or each task and cannot achieve high performance and high reliability.

There is no means to define the correlation of tasks in reliability.

Therefore, in order to realize a highly reliable distributed processing system, a mechanism for reflecting the correlation of reliability in terms of multiplexing is required in addition to the correlation of tasks on the amount of computation as in the prior art.

It is an object of the present invention to define the reliability and urgency of each operator or each task to achieve high performance and high reliability.

Another object of the present invention is to suppress the possibility of stopping the protection relay function and to assign a task so that the protection relay function can be realized with a small number of calculators.

1 is a configuration of a high reliability real-time distributed processing system to which the present invention is applied.

2 is another configuration of a highly reliable real-time distributed processing system (No. 1).

3 is another configuration of a highly reliable real time distributed processing system (No. 2).

4 is an operation (No. 1) at the time of a device failure.

5 is an operation (No. 2) at the time of a device failure.

6 is an illustration of task scheduling by a highly reliable real time distributed processing system.

7 is an illustration of a schematic diagram of a database of an executable operator control mechanism.

8 is a time sectional view (No. 1) of task scheduling.

9 is a time sectional view (No. 2) of task scheduling.

10 is a time sectional view (No. 3) of task scheduling.

11 is a design tool from an external communication network.

12 is a configuration of a distributed processing system having a channel monitoring task.

13 is a configuration of a digital protective relay device to which the present invention is applied.

14 is an illustration of a schematic diagram of a database of an execution operator control mechanism of a digital protection relay.

Fig. 15 is an illustration of a schematic diagram of a database of an input / output control mechanism of a digital protection relay.

16 is an illustration of a text form of a database of an executable operator control mechanism.

17 is an example of a text form of a database of an input / output control mechanism.

18 is an illustration (system configuration) of a digital protective relay device to which the present invention is applied.

19 is a conceptual diagram of a task in an example of a digital protective relay device.

Fig. 20 is a flowchart of operation at startup of the execution operator control mechanism in the example of the digital protective relay device.

Fig. 21 is an initial task arrangement at start-up in the example of the digital protective relay device.

Fig. 22 is a flowchart of the normal operation of the execution operator control mechanism in the example of the digital protection relay device.

Fig. 23 is an assumption of more than a calculator in the example of the digital protection relay device.

Fig. 24 is a fault tolerant operation 1 in the example of the digital protective relay device.

25 is a fault-tolerant operation (2) in the example of the digital protective relay device.

In a real-time processing device that requires reliability such as power control and traffic control such as aviation and railway, it is necessary to realize a fail safe backup by multiplexing in order to maintain normal operation against anomaly of the calculator. . When this processing purpose is to be realized by a distributed processing system, malfunction cannot be prevented only by strengthening the failure resistance of the system itself. Based on the concept of multiplexing of operations and multiplexing of algorithms, it is necessary to realize a double safety backup by providing a plurality of tasks.

In this case, it is necessary to consider the relationship of a plurality of tasks. As a correlation of tasks, computing with the same operator may not be constrained or appropriate. Therefore, in the highly reliable real-time distributed processing system of the present invention, an execution operator control mechanism for controlling such correlation is provided.

The execution operator control mechanism manages in the database about the urgency and importance of the execution time of each task, and the degree of reliability exclusivity related to the task correlation, in order to determine the operator to execute each task.

The execution operator control mechanism may refer to the regulations in the database to determine the task arrangement most suitable for the situation of the distributed processing system.

Considering the urgency and importance of each task's execution time, it maximizes CPU utilization, prioritizes the most important tasks, and selects tasks to run without compromising the degree of reliability exclusivity associated with the tasks. A process of selecting an operator to which the process and its related tasks are assigned can be performed.

In addition, if task scheduling is executed on the above-mentioned distributed processing system, high reliability can always be realized intentionally. Even in the case of failure or maintenance of the calculator, function retention is achieved to realize both CPU usage efficiency and reliability.

In addition, the present invention is an information processing apparatus for processing a plurality of tasks for realizing a protection relay function with a plurality of calculators connected by a network, comprising: means for storing a combination of tasks to be processed by a specific calculator among a plurality of tasks; Means for selecting any one of the combinations of tasks to be processed by the particular calculator, and performing the selected combination of tasks with the particular calculator.

In addition, the present invention is an information processing apparatus for processing a plurality of tasks for realizing a protection relay function with a plurality of calculators connected by a network, the means for storing a combination of tasks not processed by a specific calculator among the plurality of tasks. And means for executing a combination of tasks except for the combination of tasks not processed by the specific calculator with the specific calculator.

Hereinafter, embodiments of the present invention will be described in detail with reference to the accompanying drawings. The first embodiment is shown in FIG.

A plurality of calculators, examples of which are three calculators 102 to 104 are connected to the network via the communication network 101. Each operator here is an operator that performs digital operation. One calculator has a basic configuration such as a power supply, a CPU, and a memory independently, and is a device that is not affected by operation performance by a stop or failure of another calculator. In addition, since each operator has the same program execution environment, independent program modules can be executed by the entire operator. In the figure, box-shaped references 105-109 described inside the calculators 102-104 represent abstractly program modules to be executed inside each calculator. Each operator can execute a plurality of independent program modules simultaneously.

It is required to execute a plurality of tasks using the plurality of calculators described above. For each task, execution time is limited. In the example shown in FIG. 1, it is required to execute all five tasks 105 to 109 of tasks 1 to 5.

In the above-described plurality of tasks, the calculation by the same operator may be limited or not appropriate as the correlation between the tasks. For example, if one function is equipped to calculate the AND of the operation results of the multiplexed tasks in order to prevent malfunctions, it is not desirable to calculate these tasks with the same operator for reliability reasons. This is hereinafter referred to as intertask constraint in reliability.

Execution operator control mechanisms 110 to 112 provided in one or a plurality of processing apparatuses perform processing for assigning tasks in consideration of the rule of inter-task constraints in reliability. There are several ways to realize this.

First, as shown in Fig. 1, the execution operator control mechanisms 110 to 112 can be realized in a structure that holds a database. The execution operator control mechanism obtains in advance inter-task constraints on the reliability of each task and stores it as the database contents 113.

The database contents describe the runtime constraints of each task and the inter-task constraints on reliability. In Fig. 1, the technique 114 associated with task 1 includes a runtime evaluation index 115 relating to runtime constraints for determining real time and an execution operator evaluation index 116 relating to a relationship between task 1 and task 4. Express

According to this, in task 1, the urgency and importance of the execution time are defined by the execution time evaluation index 115, and the degree of reliability exclusivity from task 4 is defined by the execution operator evaluation index 116. . In the same way, each of the other tasks and the mutual task is described. That is, the processor is configured to obtain information for determining the urgency, importance, and reliability of the processing contents composed of the plurality of tasks from the database contents 113.

Execution operator control mechanisms 110 to 112 obtain information of the database contents 113 and determine an operator that executes each task. Considering the execution time constraints of each task, the execution operator control mechanisms 110 to 112 maximize CPU utilization, give priority to the most important tasks, and execute without compromising control conditions between tasks in terms of reliability. Processing for selecting a task (task selection processing) and processing for selecting an operator to which the considered task is assigned (operator selection processing) are performed.

In the example shown in FIG. 1, strict reliability inter-task constraints are imposed on task 1 and task 4, task 1 and task 2, task 2 and task 3, task 3 and task 5. Therefore, the task installation pattern is created so that these combinations are not installed in the same calculator. One solution is the task assignment pattern shown in FIG. Tasks 1 and 3 are assigned to operator X, tasks 2 and 4 are assigned to operator Y, and task 5 is assigned to operator Z. All of the above-described intertask tasks of reliability are satisfied.

Next, another means by which the execution operator control mechanism considers the rule of inter-task constraint on reliability is described. The execution operator control mechanism can achieve the above-described effects without having a database.

In the example shown in FIG. 2, each task has an execution operator evaluation index for each of the correlations between the execution time evaluation index for the task and other tasks in the program module. The communication network 101 and the plurality of calculators 102 to 104 are the same as those described in FIG. The difference from the embodiment shown in FIG. 1 is that tasks 201 to 205 incorporate the aforementioned indicators. Execution operator control mechanisms 206 to 208 shown in FIG. 2 are added with a mechanism for acquiring, via communication, an indicator embedded in tasks 201 to 205. Therefore, the execution operator control mechanism can perform the above-mentioned task selection process and operator selection process by referring to the indicators communicated from the task and transmitted when necessary.

Another method is described as the means by which the execution operator control mechanism considers the rules of inter-task constraints on reliability. Even if the indicators for all other tasks are not pre-built in the program module, the above-described effects can be obtained. In the example shown in FIG. 3, a method is shown in which each task holds a task name based on a predetermined grammar. The communication network 101 and the plurality of calculators 102 to 104 are the same as those described in FIG. The difference from the embodiment shown in FIG. 1 is that tasks 301-305 have a preset task name. Execution operator control mechanisms 306 to 308 shown in Fig. 3 refer to the task names of tasks 301 to 305 to calculate execution operator evaluation indexes. To do this, the execution operator control mechanisms 306 to 308 hold predetermined grammar rules 309. The grammars 310 to 312 describe exclusive constraints by the same operator as a relationship between two task name patterns.

The execution operator control mechanisms 306 to 308 recognize the task name at the installation stage of the program module. The execution operator control mechanisms 306 to 308 extract the grammar matching the task name from the grammars 310 to 312 with reference to the grammar rules 309 to obtain the inter-task constraints on reliability. In the same manner as the embodiment shown in Figs. 1 and 2, the execution operator control mechanisms 306 to 308 can perform task selection processing and operator selection processing.

As shown in Figs. 1 to 3, the execution operator control mechanism does not need to be executed by one calculator. As a function of the OS, the mechanism can be installed in the entire calculator. Master-slave configurations can be used. Furthermore, it may be installed in a calculator, for example, a remote control console, which is intermittently connected through an external communication channel instead of the communication network 101.

By introducing the above-described execution operator control mechanism, both of the constraints of real-time and reliability can be realized according to the configuration and performance of the calculator.

Next, a configuration for realizing a function of utilizing a task maintaining high reliability using the present invention will be described. In addition to the functions of the task selection process and the operator selection process of the present invention, a distributed processing system having functions of task reproduction and operator movement is assumed.

The above configuration will be described with reference to FIG. In the example shown in FIG. 4, on the basis of the example shown in FIG. 1, it is assumed that a failure occurs in the calculator X. FIG. In addition to the failure in the calculator X, the communication paths 401, the plurality of calculators 402 to 404, the tasks 405 to 409, and the execution operator control mechanisms 410 to 412 are the same as those shown in FIG.

In the example shown in this figure, tasks 1 to 5 are tasks requiring the same amount of computation. However, for the urgency and importance of the task, assume that task 5 is in the highest priority state. In addition, only four of the tasks 1 to 5 may be executed in the calculator Y and the calculator Z.

In this state, if a failure of the operator X is detected by the regular monitoring of the execution operator control mechanism 410 installed in the operator X or the execution operator control mechanisms 411 and 412 provided in the operator Y and the operator Z, the execution operator control mechanism 411 412 to redo the task selection process and the operator selection process corresponding to the failure of the calculator X.

The state of the distributed processing system after reexecution is shown in FIG. The failure calculator 502 and the normal calculators 503 and 504 are connected to the communication path 501. The task selection processing and the operator selection processing are performed by the execution operator control mechanisms 505 and 506. The assignment of the three tasks 507 to 509 (task 2, task 4, task 5) does not change from before the failure occurs. The difference from before the failure is that task 510 is newly assigned to operator 504 instead of task 1 calculated by operator 502.

This assignment is derived from a combination that realizes both the nature of the task individual and the nature of the correlation of the task. The task selection process maximizes the importance and urgency of the entire task, and the operator selection process maximizes the reliability of the combination of tasks to be installed in the same calculator. In the example shown in FIG. 5, since the computation amounts of each task are the same, a plurality of installation patterns for executing four tasks in two calculators may be considered. However, among the installation patterns, the installation pattern that satisfies the execution time evaluation and the execution operator evaluation shown in FIG.

The only ones obtained are operator Y (task 2, task 4) and operator Z (task 1, task 5).

When the execution operator control mechanism processes the task selection and the operation selection, as shown in FIG. 5, the failure detection is triggered, and the replacement task 510 of task 1 calculated by the operation unit 502 is calculated by the operation unit 504. The function of automatically assigning to can be realized.

Next, using the present invention, a configuration for realizing a function for executing task scheduling maintaining high reliability will be described. The execution operator control mechanism that executes task scheduling, in addition to the function described above, grasps the presence or absence of a task to be generated in the future from the current calculation cross section. The execution operator control mechanism performs the above-described task selection processing and arithmetic selection processing for each time section in which the task to be processed is generated, obtains an optimal solution through a plurality of time sections, and realizes task scheduling.

This embodiment will be described with reference to FIGS. 6 to 10. Fig. 6 conceptually illustrates the schedule of occurrence of tasks identified by the execution operator control mechanism and the results of task scheduling assigning these tasks to a calculator. In the figure, the horizontal axis 601 represents a time axis indicating the occurrence of a task related to task 1 and its processing. An arrow 602 below the horizontal axis 601 indicates the occurrence of a request to execute task 1. The rectangle 603 above the horizontal axis 601 represents the execution of task 1. The operation name inside the rectangle 603 represents an operation responsible for execution. In this case, the task 1 execution request 602 is assigned to the execution 603 by the calculator X. In the execution 603, when an execution corresponding to the new execution request 604 is allocated, the operation may appear as the long direction 605. FIG.

In the first time section 606, the above processing can be realized by task assignment by the execution operator control mechanism described above. The task request generated in the time section 604 has the same contents as in the first embodiment. That is, the task assignment configuration shown in FIG. 1 is used. That is, the assignment of the calculators of the tasks 1 to 5 shown in FIG. 6 is consistent with the contents of FIG. 1.

In the second time section 607, a new task 1 request is generated. In response to this request, task scheduling by the execution operator control mechanism maintains the reliability inter-task constraints shown in FIG. 7 to newly plan task assignment in the second time section. In addition to the task assignment state shown in FIG. 1, task 1 is newly assigned to operator Z, such as execution 605. The result is the state shown in FIG. In Fig. 8, in the operator Z 801, there is no inter-task constraint on the reliability between the newly assigned task 802 (the task shown by the thick line in the figure) and the task 803 being executed. In total, the inter-task constraints of reliability shown in FIG. 7 are satisfied.

At the time point of the third time section 608, the task 1 execution process 603 executed by the calculator X and the task 2 execution process 603 executed by the calculator Y are ended, and conversely, the task 5 execution request 611 is performed. ) Is received. Task scheduling by the execution operator control mechanism executes the task assignment shown in FIG. From the viewpoint of the processing performance of the calculator, the calculator capable of executing task 5 is the operator X 901 or the operator Y 902. However, if task 5 is assigned to operator X, considering task 903 (task 3) being executed, the inter-task constraints in reliability are not followed. As an assignment result for obtaining the best evaluation, execution of the new task 904 (task 5) by the operator Y is selected. This is because there is no problem imposed on the reliability with the task 905 (task 4) being executed.

At the time point of the fourth time section 609, all execution processing of tasks other than the task 5 execution processing 612 executed by the calculator Y ends. Here, the execution request 613 of task 1 to task 5 is received. By the same processing as described above, task scheduling by the execution operator control mechanism executes the task assignment shown in FIG. In addition to the task 1002 (task 5) being executed by the calculator Y 1001, five tasks 1003 to 1007 are newly allocated. There is no violation of the inter-task constraints on reliability.

As described above, in each time section in which an execution request occurs, task scheduling can be realized when the task selection and the operator selection are processed together with the currently executing task.

In the method for performing task scheduling by assigning tasks in a plurality of time sections, a plurality of methods may be considered depending on the difference in the method of installing the optimization means. In the example described above, task scheduling is performed sequentially along the flow of the time section. Apart from this, task scheduling can be executed by optimizing the task assignment in the entire time section simultaneously to solve the combination problem of minimizing the difference in task assignment status between adjacent time sections.

By this task scheduling function, the distributed processing system can always realize the planned high reliability. For example, even when a start request for a new task other than the task identified by the execution operator control mechanism occurs in the system or externally received a start command for a new task, task scheduling is re-executed with the reflected change, thereby performing High reliability can be maintained. In addition, in the failure and maintenance of the calculator, even when a command for the utilization state of the calculator is received via the external calculator, the task scheduling is re-executed after the utilization state of the calculator is assumed according to the instruction contents. The function of realizing both use efficiency and reliability can be maintained.

Next, as a second embodiment, a development tool to which the present invention is applied is shown in FIG.

The development target processing system is the same as the embodiment shown in FIG. Through a communication path 1101 used for control, a plurality of calculators, that is, three calculators 1102 to 1104 in this example, are connected to a network. These calculators are required to execute a plurality of tasks 1105 to 1109 whose execution time is limited.

In order to design such a processing system, an external development calculator is used. The development calculator 1111 is connected via an external communication network 1110 which is not used for control. By using such a development calculator, the basic design of the above-described task and various setting values are set.

The development calculator 1111 has an execution operator control mechanism 1112 having the same function as described in the embodiment shown in FIG. Therefore, by instructing the device configuration of the development target processing systems 1101 to 1104, initial task assignment can be performed to realize both CPU usage efficiency and reliability.

By using such a development tool, an initial design for performing the selection process of the task to be executed and the selection process of the calculator for calculating the selected task can be performed, and a highly reliable real time distributed processing system can be designed. In addition, tasks can be added, deleted, or modified through an external communication network, thereby realizing most of the functions maintained by the highly reliable real-time distributed processing system without adding an execution operator control mechanism to the computing operator constituting the processing system. .

As a third embodiment, a distributed processing system having a function of monitoring a communication path will be described below utilizing the present invention.

In order to configure the processing system, a plurality of calculators, that is, three calculators 1202 to 1204 in this example, are connected to the network through the communication path 1201.

Here, a channel monitoring task for verifying the communication path is added, making the processing system more reliable. As a communication channel monitoring method, a plurality of tasks communicating through a monitored communication channel are formed as a set, and the validity of the mutual communication contents is verified, thereby monitoring the normality of the communication path.

In this case, the assignment and task scheduling of these monitoring tasks presents a complex problem. First, in order to realize the monitoring function, it is preferable that not only the minimum required monitoring task but also the monitoring task to be executed are executed as much as possible. However, task execution, which is the original purpose of processing, should not be interrupted. Second, the individual tasks that make up the set must be executed by different operators than the tasks of the same set. Allocation and scheduling of such monitoring tasks with a plurality of constraints can be realized using the present invention.

In the same manner as the example described in the embodiment shown in Fig. 1, the plurality of calculators in the processing system are provided with execution operator control mechanisms 1205-1207. These execution operator control mechanisms control the meaning of the task to be executed as the database contents 1208. The database contents describe constraints on execution time of each task and constraints between tasks on reliability. In FIG. 12, the technique 1209 related to task 1 includes execution time evaluation indicators 1210 related to execution constraints for determining real-time and execution operator evaluation indicators 1211 related to the relationship between task 1 and task 4. Express

In this example, it is assumed that there are two differences compared with the embodiment shown in FIG. First, the execution time evaluation index 1213 related to the channel monitoring task A 1212, the execution time evaluation index 1215 related to the channel monitoring task B 1214 and the execution operator evaluation index 1216 between tasks are provided. Suppose it is added.

Secondly, the execution time evaluation index 1217 related to task 5 is set to have a lower priority than the execution time evaluation indicators of the task 1 to task 4 and the execution time evaluation index of the monitoring task. In other words, task 5 is set to not be a required task.

Here, two tasks are formed as a set, and the function of monitoring a communication path is added. Since the watchdog function is required at the minimum, a higher execution priority than task 5 is set.

Given this setting to the execution operator control mechanisms 1205 to 1207, task assignments 1220 to 1225 are executed for the three operators 1202 to 1204. Execution of task 5 is omitted. In addition, for the other six tasks, the arrangement of each task is determined to satisfy the mutually exclusive relationship.

When the tasks for monitoring the communication paths are set as part of the control object of the execution operator control mechanism, a highly reliable real time distributed processing system for grasping the normality of the communication paths can be realized.

Furthermore, by utilizing task scheduling using an executable operator control mechanism, the communication path can be periodically monitored. By periodically requiring the processing of a set of monitoring tasks with a lower execution priority, as high a reliability as possible can be realized within the range of excessive computing power of the processing system.

In the above-described embodiment, monitoring of the communication path is described. By the same method, the monitoring function of the calculator can be installed. First, as an operator monitoring task, a test for checking a predetermined expression is executed, and a plurality of tasks for verifying a result through communication with a plurality of other calculators connected by a network are prepared. Then, for the execution operator control mechanism, a constraint or condition is given that each operator monitoring task is executed by a different operator. By doing so, the operator monitoring task is automatically subjected to proper task assignment and scheduling, and detection of abnormality of the operator or failure of the network can be realized.

As a fourth embodiment, consider an example in which a digital protection relay device for protecting one or a plurality of power devices is realized using a processing system to which the present invention is applied. This is shown in FIG.

First, as an input target device, voltage and current information 1302 used for an accident detection function required for device protection and an observation function required for various controls for each of the power target devices 1301 to be protected and controlled, such as a bus, a transformer, and a transmission line. Is transmitted to an input device 1303 connected thereto. In the same manner, for the control target power device 1304 such as a breaker as the output target device, a control signal 1305 such as a shutoff command is transmitted to the output device 1306 connected to each device.

These input devices 1303 and output devices 1306 are connected to the plurality of calculators 1308 to 1311 through a communication path 1307. In this case, the calculator is a calculator that performs digital calculation. One calculator is a device which is provided with basic components such as a power supply, a CPU, and a memory independently, and does not affect the computing performance by stopping or malfunctioning other calculators. In addition, since each calculator has the same program execution environment, independent program modules can be executed by the entire calculator. In the figure, box-shaped reference numerals 1311 described inside the calculators 1308 to 1310 abstractly represent a program module to be executed inside each calculator. Each operator can execute one or a plurality of independent program modules simultaneously.

The assignment of the execution of each program module to the operator is controlled by the execution operator control mechanism 1312. This is the same function as the execution operator control mechanism of the first embodiment shown in Fig. 1 described above. Therefore, the database inside the execution control device describes the execution time constraints of each task and the inter-task constraints on reliability.

In FIG. 13, the execution operator control mechanism controls the database contents shown in FIG. Execution constraints on the execution time and reliability of each task are described. In the technique shown in Fig. 14, the technique associated with each task includes a runtime evaluation index 1402 relating to runtime constraints for determining real time and an execution operator evaluation index 1403 relating to other tasks. .

For example, among the calculators protecting the device A, the program modules "device A, the main system, the normal system" that always execute the main protection operation, and the "waiting to be immediately switched and backed up when the program module becomes inoperable" Device A, Main, Standby System "must be implemented by different operators. Therefore, the inter-task constraint on reliability is imposed on the relationship between the task "device A, main, clock" and the task "device A, main, atmospheric system".

In the same way, in order to prevent malfunctions, "Device A, Main, Clock" and "Device A, fail safe, Clock" must be executed by different calculators. Therefore, the constraint between tasks on reliability is imposed on the relationship of two tasks. When such an operator constraint is imposed, the possibility of malfunction of the protection function for the device A does not occur if two calculators on which the program modules representing the two tasks are executed are not inoperable at the same time.

In addition, even if a single failure of the device occurs, task assignments that maintain high reliability may be further imposed. For example, even when a task of the clock is switched to the standby system due to a device failure, if the main task and the double safety task are designed to be executed by different calculators, the "device A, main, clock" and In the task relationship of " device A, dual safety, standby system ", a task execution operator evaluation index can be set as an inter-task condition on reliability. In this case, the exclusive condition 1404 may be set to a priority lower than the inter-task constraint 1403 on reliability.

As described above, the design shown in FIG. 14 is executed for all tasks related to the device A. FIG. In addition, according to the same concept, a task related to the device B is designed. By the execution operator control mechanism provided with such a database, an operator for executing each program module is appropriately determined. As a result, the task assignment shown in Fig. 13 is obtained. The number of operators cannot satisfy the constraints and imposed conditions between tasks on overall reliability. An arrangement is set which satisfies the high priority constraint 1403 and at the same time satisfies the exclusive condition 1404 as much as possible. At any point in time, a system with high reliability can be realized by utilizing the processing power of the entire system.

Furthermore, in the above embodiment, the control mechanism is also provided to control the above-described input / output relationship. As shown in Fig. 13, the input / output devices J to M are fixedly connected to the devices A to D. The control mechanism is configured to transmit and receive input and output information of the device via a multi-drop communication network. The input / output relation control mechanism 1313 manages a situation in which a task in the calculator transmits a control signal to an arbitrary output device using the V and I information of the arbitrary input device.

As an example, consider the operation on the system shown in FIG. 13 in which the input device J 1314 and the input device J '1315 are prepared as input devices connected to the device A. As shown in FIG. The input / output relation control mechanism receives a designation 1503 in which the input device J 1501 and the input device J '1502 are in mutually interchangeable relationship by the database contents schematically shown in FIG. Therefore, upon detecting a failure of the input device J, the input / output relation control mechanism instructs each calculator to replace the input data to be transmitted by the input device J with the input data to be transmitted by the input device J '. By doing so, the influence of the failure of the input device J can be avoided.

In the above description, the contents of the database shown in Figs. 14 and 15 are specified and installed on a rule base at the design stage of the processing system or during its maintenance. This method is described with reference to the examples.

16 shows the data format to be given to the execution operator control mechanism. The contents schematically represented in FIG. 14 may be given as text data according to a predetermined grammar.

In Fig. 16, one line 1601 indicates a comment sentence. Below 2 rows 1602, protection functions associated with device A and device B are set.

First, a task related to protection of the device A is declared. By the description in the third row 1603, a program module that protects the functions of the device A, the main, and the clock is declared under the name of "module_A1" as a task that always performs processing. The string "Device A, Main, Clock" below the mark # is a comment. Further, by the description in the fourth row 1604, a program module for processing the functions of the device A, the main, and the standby system is declared under the name of "module_A2" as a task for performing the standby processing. In the same way, for device A, a total of four tasks are declared. Techniques related to the real time of each task, for example, task processing start time, end time and priority, are set separately.

Next, the relationship between the clock and the atmospheric system related to the protection of the device A is declared. In line 7605, it is defined that module_A1 is backed up by module_A2. Hereinafter, a backup of the double safety function is described in the same manner.

At the end of the definition data of the device A, an exclusive relationship between tasks is declared. In line 9606, the constraint that the "Device A, Main, Clock" function and the "Device A, Dual Safety, Clock" function are assigned to different calculators is defined as the highest priority. In 11 lines 1607, it is defined as a lower priority condition that it is preferable to assign the "device A, main, clock" function and the "device A, dual safety, clock" function to different calculators. In the same way, five relationships existing between four tasks are defined below.

In the text data given to the execution operator control mechanism shown in Fig. 16, the setting relating to the device B is also described in the same manner as the protection relating to the device A. In this way, a rule of reliability inter-task constraints schematically shown in FIG. 14 can be obtained.

Next, the data format to be given to the input / output relation control mechanism will be described. Contents schematically represented in FIG. 15 may be given as text data shown in FIG. 17 according to a predetermined grammar.

In Fig. 17, one line 1701 indicates a comment sentence. Below two lines 1702, settings relating to an input device, settings relating to an output device, and settings relating to input and output are executed.

First, the input amount to be transmitted by the input device J is declared. By the description in the third row 1703, the name of "a_bus_v1" is set as the first input amount of the input device J. Hereinafter, the total amount of input transmitted by the input device J is declared and its names are set.

In the same way, the amount of output received by the output device is also declared. The output device L is defined below the 17th line 1704. In the same manner as the description relating to the input quantity, an output quantity is declared and the name is set in 18 lines 1705.

Finally, for each task, the input / output amount used by the task is set. Below 22 lines 1706, the input / output amount of the task module_A1 is set. Arguments held by the program module for realizing the above-described task are set sequentially. In line 23 1707, it is defined to use the input a_bus_v1 as the first argument. In the same manner, all the necessary inputs and outputs are defined by the above-described program module. Then, it is set that can be substituted for the input / output amount used by the task. In line 1708, when an error occurs in the input a_bus_v1 used as the first argument, it is defined to use a2_bus_v1 instead.

By the text data given to the input / output relationship control mechanism shown in Fig. 17, a rule relating to the relationship between the input / output amount schematically shown in Fig. 15 and its replacement can be given.

By introducing the expression method based on the grammar shown in Figs. 16 and 17, designers and operators can write rules directly in the text. In addition, by a mechanism for outputting the text file of the representation from a design and operation tool with a GUI, designers and operators obtain an easier environment for design and operation.

As described above, if the highly reliable real time distributed processing system having the execution operator control mechanism shown in Embodiment 1 controls the task of realizing the digital protection relay of the power equipment, the highly reliable protection relay system can be realized.

There are several correlations between the tasks to be controlled. For example, as described in the above-described Embodiment 3, as a correlation between the upper clock task for the first calculation purpose and the waiting system task, the constraints or conditions of execution by different calculators can be cited. According to this, a digital protection relay is realized which prevents the protection relay function from being stopped in the event of a calculator failure.

In addition, a protection relay including a main detection task for realizing the main detection function of the protection relay and an accident detection task for realizing the accident detection function as a double safety element has a correlation between the main detection task and the accident detection task of the same protection relay. As constraints or conditions of execution by different operators may be cited. According to this, a digital protection relay for preventing malfunction is realized.

In addition, a protection bus including a package bus protection task for realizing a package bus protection function of a bus protection relay and a split bus protection task for realizing a split bus protection function includes a package bus protection task and a split bus protection for the same bus. As a correlation between tasks, constraints or conditions of execution may be cited by different operators. According to this, a digital protection relay for preventing a bus protection relay malfunction is realized.

Furthermore, for each of the transmission lines composed of parallel lines, a protection relay having a transmission line task for realizing the transmission line protection relay function is restricted in execution by different calculators of a plurality of transmission line tasks for transmission line protection belonging to the same parallel line. Or conditions may be cited. According to this, a digital protection relay is realized which minimizes the effect of a calculator failure on the protection of parallel transmission lines.

As described above, a highly reliable protection relay system controlled by a highly reliable real time distributed processing system will be described as a fifth embodiment with reference to the drawings.

First, the entire system image is shown in FIG. A plurality of calculators, for example, three calculators 1802 to 1804 are connected to a network through an optical communication path 1801 for realizing communication of an optical medium. Each calculator shown here is a calculator that performs digital calculations. One calculator has a basic configuration such as a power supply, a CPU, and a memory independently, and is a device that is not affected by operation performance by a stop or failure of another calculator. In addition, since each operator has the same program execution environment, independent program modules can be executed by the entire operator. In addition, each operator obtains a signal synchronized with a commercial frequency through the optical communication network 1801. Therefore, the protection relay operation synchronized with the electric angle of the commercial frequency can be performed.

The optical communication network 1801 is connected to an input / output device as well as a calculator. This form is shown in the figure taking into account one phase of the power line. A breaker 1806 and a transformer 1807 are installed in the power line 1805 for supplying commercial power, and each of the devices includes input and output devices 1808 to 1811. The system has a form in which these input / output devices are connected to the optical communication network 1801.

The input / output devices 1808 to 1811 have an analog-to-digital conversion function, a communication function, and an electro-optical conversion function, and are mounted to the breaker 1806 and the transformer 1807. The input / output devices 1808 and 1809 are dual communication terminals of the breaker 1806 and realize communication functions for voltage-current information from the breaker to the calculator and cutoff control information from the calculator to the breaker. In the same manner, the input / output devices 1810 and 1811 realize a communication function as a dual communication end of the transformer 1807. Each input and output device and operator are independent of each other in terms of isolation as well as in terms of power supply.

In the same manner as the embodiment described above, in such a system, an execution operator control mechanism 1812 and an input / output relation control mechanism 1813 are installed in one or a plurality of arbitrary calculators.

In a highly reliable protection relay system controlled by a highly reliable real time distributed processing system having the above-described configuration, processing of a plurality of tasks with real time is required. Such tasks are represented abstractly in FIG. 19. Boxed reference numerals 1901 to 1906 shown in FIG. 19 indicate tasks processed by the calculators 1802 to 1804.

The task indicated by reference numeral 1901 is hereinafter referred to as "Task1" or "TrPrMN". Task1 is a task for realizing the main transformer differential relay element 87T operation and the sequence operation that defines the firing condition of the element.

The task indicated by reference numeral 1902 is hereinafter referred to as "Task2" or "TrPrFD". Task2 is a task for realizing a different algorithm accident detection function that performs the double safety function of Task1 described above and a sequence operation that defines the firing condition of the element.

The task indicated by reference numeral 1903 is hereinafter referred to as "Task3" or "TrPrMN-FT". Task3 is a task for realizing the standby function which performs the fault tolerance function of Task1 mentioned above. If Task1 is inoperative, in order to immediately back up its functionality, Task3 performs the least amount of operations of the same algorithm as Task1 and continues to wait. Task3 performs an operation that continues processing required to maintain a prioring value such as reading of an input amount or digital filtering.

The task indicated by reference numeral 1904 is hereinafter referred to as "Task4" or "TrPrFD-FT". Task4 is a task for realizing the standby function of performing the above fault tolerance function of Task2 in the same manner as Task3.

The task indicated by reference numeral 1905 is hereinafter referred to as "Task5" or "SS-P / P '". Task5 performs the function of monitoring the normality of the input / output device P (1808 shown in FIG. 18) and the input / output device P '(1809 shown in FIG. 18) described in FIG. For example, a hardware function is added to the input / output device that superimposes the twelfth harmonic component with respect to the commercial frequency as an inspection signal on the amount of electricity (analog value) input from the power line. By doing so, when an arbitrary operator monitors the test harmonic component included in the information transmitted or received from the input / output device, the normality of the input signal can be monitored.

The task indicated by reference numeral 1906 is hereinafter referred to as "Task6" or "SS-Q / Q '". Task6 performs the function of monitoring the normality of the input / output device Q (1810 shown in FIG. 18) and the input / output device Q '(1811 shown in FIG. 18) described in FIG. 18 in the same manner as Task5.

Real time is required for these multiple tasks. In the example shown in FIG. 19, it is required to perform six tasks of Task1 to Task6 within a predetermined execution time using the three calculators 1802 to 1804 shown in FIG.

In FIG. 19, the predetermined runtime constraints for each of the six tasks of Tasks 1-6 are shown 1907-1912. In the same manner as the above-described embodiment, the data relating to these runtime constraints is given in advance to a database controlled by the execution operator control mechanism. The contents of each runtime constraint will be described below.

Shown in the runtime constraint 1907 is that Task1 has an execution priority of Level 1, the request for the task occurs at a time frequency of 30 deg, and the deadline from generation to processing is 30 deg. For Level 1, Level 1 means the highest priority task, because smaller values indicate higher priority. The processing time period deg is expressed in units of an electric angle for one cycle of a commercial frequency, and designates a time when performing a protective relay operation synchronous with the electric angle of a commercial frequency as described in FIG. 18.

In the same way, the runtime constraint 1908 indicates that Task2 has an execution priority of Level 1, the frequency of occurrence is 30 deg, and the processing time is 30 deg. In other words, it is defined that Task1 and Task2 which implement the main protection function are the highest priority tasks.

In addition, the execution time constraints 1909 and 1910 are priority tasks in which Task 3 and Task 4, which realize fault tolerance of protection functions, are the highest priority, and in the same manner, the frequency of occurrence is 30 deg and the processing time is 30 deg. To stipulate.

On the other hand, the execution time constraints 1911 and 1912 have the lowest execution priority by design. The frequency of occurrence is 10 sec, and the processing can be completed within the time limit of 2 sec. In addition, for Task6, a non-periodic task generation rule is also described so that the task does not occur for one minute after system startup.

In this embodiment, in addition to the execution time constraints defined for each of the above-described tasks, the reliability inter-task constraints that define the correlation of the tasks in the same manner as in the previous embodiment of the highly reliable real-time distributed processing system. This high reliability protection relay system is required.

Exclusion constraints 1913 to 1916 represent inter-task constraints in reliability that depend on the relationship of Task 1 to Task 4. In the same manner as the above-described embodiment, the data relating to these inter-task constraints on reliability are given in advance to a database controlled by the execution operator control mechanism. The contents of the constraints between the tasks in the reliability will be described below.

Exemplary content shown in exclusive constraints 1913 indicates that Task1 and Task3 have an exclusive priority of Level 1. For Level 1, since smaller values represent higher priorities, Level 1 means that the exclusive placement of the operator is considered top priority. The reason is that Task3, which realizes the fault tolerance function of Task1, is completely meaningless unless it is handled by different operators.

In the same way, the exclusion constraint 1914 specifies that, for Task2 and Task4, the exclusive placement of the calculator is considered first.

Exclusive constraints 1915 indicate that Task1 and Task2 are in exclusive priority Level 2. Tasks that realize the double safety feature of Task1 should, in principle, be handled by different operators. However, since Task2 is composed of different algorithms as the accident detection system, even if it is processed by the same calculator as the main detection Task1, it is not meaningless. Thus, inter-task constraints on reliability are defined to take into account the exclusive placement of the operator as much as possible.

In the same way, exclusive constraints 1916 define inter-task constraints on reliability to take into account the exclusive placement of the calculator as much as possible for Task3 and Task4.

Based on the above-mentioned contents of the execution time constraint and the inter-task constraint on reliability, the execution operator control mechanism determines the placement of the task and controls the execution instruction. The processing steps will be described sequentially. First, the operation at system startup, that is, the initial operation of the present embodiment will be described in detail.

20 is a flowchart showing the operation of the execution operator control mechanism at system startup. First, in process 2001, the execution operator control mechanism initializes all the operators. Then, a process 2002 is performed to grasp the state of the processing capability of each computing unit and to grasp the computing capability of each computing unit. Next, in process 2003, the execution operator control mechanism calls the specified contents relating to the entire task from the database. Specifically, the above-described runtime constraints and inter-task constraints on reliability. Among these, a process 2004 for selecting a task having a start request from the start of system startup is performed. In the example shown in FIG. 19, Task 1 to Task 5 are initial start request tasks but not Task 6.

By the process 2004, grasp of the execution time constraints related to the initial start request task and the constraints between the tasks on the reliability is completed. Based on the data relating to the constraints, in process 2005, the execution control mechanism can minimize the amount of violations, that is, follow the runtime constraints of each task, maximize CPU utilization, and prioritize important tasks. And preventing the inter-task constraints on reliability from being broken, thereby selecting a task to be implemented and an operator to which the task is assigned. For that purpose, arithmetic processing to solve the combinatorial optimization problem is performed, and task assignment is determined from these results. When the arrangement position of each task is determined, the execution operator control mechanism outputs a process of allocating a program module (hereinafter, referred to as a task module) of the task in an execution form (2006) and an instruction to execute the assigned task module. Process 2007 is performed.

Further, the execution operator control mechanism allocates the task module by precomputing not only the start of the initial task but also the task arrangement in the next time section. The next time section is the time section in which the task occurrence request is next. As shown in Fig. 19, the next time cross section of the present embodiment has a minimum task frequency of 30 deg, so that it is 30 deg later in synchronization with the electric angle from the start time.

In the process 2008, the execution operator control mechanism grasps the task request in the next time section from the real time of the task defined in the execution time constraint described above, and the next task batch optimization process 2009 in the same manner as described above. Is done. The arrangement of the assigned initial task module is compared with the arrangement of the next task module which is now determined, and a process 2010 is performed to assign the next task module to the destination calculator as necessary. In this way, there is no task generated due to external factors in the next time section, and if the task request is generated as scheduled, the task can be executed in an optimal arrangement.

Fig. 21 shows the state of the initial task arrangement allocated by the startup operation of the above-described execution operator control mechanism. Task1 (2102 in the drawing) and Task5 (2103 in the drawing) are arranged in the calculator X (2101 in the drawing), Task2 (2105 in the drawing) and Task3 (2106 in the drawing) are arranged in the calculator Y (2104 in the drawing), Task 4 (2108 in the drawing) is arranged on the calculator Z (2107 in the drawing). In this arrangement, all of the inter-task constraints (here, exclusive constraints of two tasks) on reliability, shown at 1913-1916 in FIG. 19, are satisfied. The processing of the entire task required at this stage is executed, and the whole system is in a healthy situation. For example, even if a Task6 request occurs over time, the executor operator can automatically assign Task6 to the batch portion of Task5, and plan the process of returning it to Task5 after completion, thus keeping the whole system healthy. To be maintained.

If a situation other than the above plan occurs as a whole of the system, for example, if any computing device becomes unavailable, the execution operator control mechanism should reassign the task. The processing flow is shown in FIG.

First, the execution operator control mechanism starts the above process when the estimated time (30 deg in this embodiment) elapses from the last start of the execution operator control mechanism itself. It is provided by the process 2201 for setting the signal synchronized with the electric angle as an interruption trigger.

After the start-up, the execution operator control mechanism first performs processing 2202 to determine whether or not the arithmetic capability table, that is, the state of the arithmetic operator is changed by collecting various monitoring results. If a device abnormality is found, the process of performing the fault tolerance function is started. This will be described later. Hereinafter, the process when there is no apparatus abnormality is demonstrated.

First, in process 2203, the execution operator control mechanism reads the current task request. Any task of this system must present a task request to the execution operator control mechanism when it is necessary to start another task as a result of its operation. For example, the main detection task of Task1 has a function of outputting an abnormal input amount, and is configured to present a request for precise abnormality monitoring at that time. That is, Task1 sends a Task6 request to the execution operator control mechanism under the condition of detecting abnormal input / output. This request is read by process 2203.

Whether there is a difference between the present task request and the scheduled task of the next section in the previous execution of the execution operator control mechanism is judged by the determination process 2204 of the existence of the unscheduled task request, and the process branches.

If there are no unscheduled tasks, the task module executing in this time step is previously assigned to the previous execution of the execution operator control mechanism. Thus, the current task execution command is immediately sent by the process 2205.

If there is an unscheduled task, the process 2206 reoptimizes the placement of the executable task, that is, the task module of the task group assigned to the calculator. If the execution time constraint of an unscheduled task is a high priority, the execution operator control mechanism may relocate the tasks. Whether they are actually relocated depends on constraints between tasks in terms of reliability with other assigned tasks.

In either case, the executive operator control mechanism automatically selects the optimal placement within the executable task. Processing 2207 then sends an execution command to the optimal batch task within the executable task described above. If there is a time allowance compared to the task module assignment, the process 2208 allocates some tasks as supplementary tasks, and the process 2209 sends an immediate execution command to the supplementary tasks.

Even if there is an unscheduled task or no unscheduled task, the processes 2210 to 2212 perform batch optimization processing for the task request in the next time section, and perform the process of assigning the next task module to the destination calculator as necessary. do. This process is exactly the same as in 2008-2010 described in FIG.

Next, the case where the device abnormality is recognized by the determination process 2202 and the fault tolerance function is performed will be described.

First, the process 2213 searches for the existence of a waiting system task (hereinafter referred to as a fault tolerance task) that performs a fault tolerance function on a task executed in the apparatus in which an abnormality occurs. In the example shown in FIG. 19, for [Task1] TrPrMN, it is specified that the [Task3] TrPrMN-FT is a fault tolerant task. As described above, the fault-tolerant task for the task running in the apparatus in case of an abnormality is retrieved from the database.

In this case, if a fault tolerance task exists, it is necessary to immediately switch the corresponding task from the standby processing mode to the main processing mode. This is realized by sending a master / slave mode switch command to the fault-tolerant task retrieved by the processes 2213 and 2214. By switching the master / slave mode, the function is immediately maintained.

After switching the master / slave mode, the task is relocated. This concept is basically the same as the handling for the occurrence of an out-of-date task already described. In a combination of already deployed task modules, the instrument selects an optimal execution module. However, if an arithmetic abnormality has already occurred, first, in process 2215, the arithmetic capability table is re-created to update the state of the arithmetic operator grasped by the execution operator control mechanism. Then, the processes 2206 to 2209 described above select the optimal task executable in the current situation, and send an execution command to it. Further, task placement optimization in the next time section is also performed by the processes 2210 to 2212 described above.

According to the above-described process flow chart of FIG. 22, when an arithmetic abnormality occurs, the fault tolerance function is immediately performed, and after the next time section, that is, after 30 deg steps, an optimal task arrangement is autonomously obtained and execution continues. .

Hereinafter, the operation | movement at the time of an operator abnormality occurrence is demonstrated using the example of the task arrangement | movement shown in FIG. 21 with reference to FIG. The task specification contents (time constraints and inter-task constraints on reliability) are the same as those shown in FIG. 19, and consider a case where an operator abnormality occurs in the same time section as the task arrangement is shown in FIG.

In Fig. 23, the calculators 2301 to 2303 are the same as the calculators 2101, 2104, and 2107 shown in Fig. 21, and the task arrangement is the same. When the operator 2301 is disabled, [Task1] TrPrMN (shown as 2304 in the figure) and [Task5] input / output P / P 'monitoring (shown as 2305 in the figure) are disabled. The operation of the execution operator control mechanism at this time will be described below according to the flowchart shown in FIG.

First, the execution operator control mechanism that interrupts and starts in the process 2201 determines the abnormality of the operator X (shown as 2301 in FIG. 23) in the process 2202. In this case, the program immediately proceeds to the fault tolerance start process. In process 2213, the instrument knows that the fault tolerant task [Task3] TrPrMN-FT is executed by operator Y (shown as 2302 in FIG. 23) for [Task1] TrPrMN. [Task5] For I / O P / P 'monitoring, there is no fault tolerance task. Thereafter, in immediate processing 2214, the instrument instructs [Task3] TrPrMN-FT to switch the master / slave mode. As shown in FIG. 24, [Task3] TrPrMN-FT 2401 shown in the figure continues to provide the functionality of Task1 instead of [Task1] TrPrMN or equivalent.

Next, at process 2215, after reflecting the situation where operator X cannot be used, the mechanism, if any, performs process 2216 by which the assigned task module plans a better task execution form. Here, specifically, the apparatus is executed by a constraint violation (exclusive Level 2) between tasks in reliability of [Task1] TrPrMN 2401 shown in the drawings and [Task2] TrPrFD 2402 shown in the drawings, and any operator. [Task3] Although a search is performed for a combination that solves the runtime constraint violation (execution priority level 2) of TrPrMN-FT, the solution does not exist in this example. Accordingly, the task arrangement form shown in FIG. 24 is an emergency arrangement for the abnormality of the calculator X. FIG. In process 2207, execution instructions are sent to tasks 2401-2403. Since there are no tasks that can be supplemented and executed, processes 2208 and 2209 do not execute task assignment and execution command issues.

The instrument then plans the optimal placement in the next time section, ie in the future section after 30 deg of electrical angle, and assigns a task module. First, in process 2210, the mechanism examines the task request. In this case, based on the task specification shown in Fig. 19, the mechanism grasps the task request of Task1 to Task4. Then, in process 2211, the mechanism performs next task batch optimization. As a result, the instrument can obtain the solution (task arrangement) shown in FIG.

When the task arrangement shown in FIG. 25 is compared with that shown in FIG. 24, the arrangement of [Task1] TrPrMN 2501 in the figure does not change, but the [Task4] TrPrFD-FT 2502 shown in the figure and the figure shown in the figure. [Task2] The arrangement of TrPrFD 2503 is replaced. Accordingly, the inter-task constraint violation (exclusive level 2) on the reliability of [Task1] TrPrMN and [Task2] TrPrFD can be solved. However, for a runtime constraint violation (execution priority level 2) of [Task3] TrPrFD-FT, no batch is found to solve this. Thus, the task arrangement shown in FIG. 25 is obtained.

Finally, the instrument allocates a next task module in process 2212. Specifically, the mechanism assigns the task module of [Task4] TrPrFD-FT to operator Y and the task module of [Task2] TrPrFD to operator Z. By doing so, the task module necessary for realizing the task arrangement shown in Fig. 25 is prepared in the next time section.

If there is no change in the operator state and the task request state in the next time section, that is, in the future section after 30 deg of the electric angle, the execution operator control mechanism will continue to perform the optimal task processing only by sending an execution command to the tasks 2501 to 2503. Can be.

According to the distributed processing system of the present invention, task assignment for realizing both real-time and reliability is executed according to the configuration and capability of the calculator. Further, by executing task scheduling, the distributed processing system can always realize high reliability intentionally. In addition, it is possible to obtain a function maintenance that achieves both CPU usage efficiency and reliability for processing in which a task start request frequently occurs, or even in the case of a failure or maintenance of the calculator.

Claims (24)

An information processing apparatus for processing a plurality of tasks for realizing a protection relay function with a plurality of calculators connected by a network, Means for storing information defining a combination of tasks to be processed by a particular operator among the plurality of tasks; Means for selecting any one of said combinations of tasks to be processed by said particular operator in accordance with said combination defined; And And means for executing the task of the selected combination by the specific operator. The method of claim 1, The plurality of tasks include a phase clock task that always performs a process included in a protection relay and a standby system task that performs the process included in the protection relay in an emergency. And the combination of the tasks that can be processed by the particular calculator does not include a combination of the clock clock task and the standby system task. The method of claim 1, The plurality of tasks include a main detection task for realizing the main detection function of the protection relay and an accident detection task for realizing the accident detection function of the protection relay, And said combination of said tasks that can be processed by said particular calculator does not include a main detection task and an accident detection task of the same protection relay. The method of claim 1, The plurality of tasks include a package bus protection task for realizing a package bus protection function of a bus protection relay and a split bus protection task for realizing a split bus protection function of a protection relay, And said combination of said tasks that can be processed by said particular operator does not include a package bus protection task for the same bus and a split bus protection task for the same bus. The method of claim 1, The plurality of tasks includes a power transmission line task for realizing a power line protection relay function of each power transmission line composed of parallel lines, And said combination of said tasks that can be processed by said particular operator does not include a plurality of transmission line tasks for transmission line protection belonging to the same parallel line. An information processing apparatus for processing a plurality of tasks for realizing a protection relay function with a plurality of calculators connected by a network, Means for storing a combination of tasks among the plurality of tasks that are not processed by a particular operator; And means for executing a combination of tasks other than the combination of the tasks not processed by the specific calculator by the specific calculator. The method of claim 6, The plurality of tasks include a phase clock task that always performs a process included in a protection relay and a standby system task that performs the process included in the protection relay in an emergency. And said combination of said tasks not processed by said specific calculator comprises a combination of said clock clock task and said standby system task. The method of claim 6, The plurality of tasks include a main detection task for realizing the main detection function of the protection relay and an accident detection task for realizing the accident detection function of the protection relay, And said combination of said tasks not processed by said specific calculator comprises a main detection task and an accident detection task of the same protection relay. The method of claim 6, The plurality of tasks include a package bus protection task for realizing a package bus protection function of a bus protection relay and a split bus protection task for realizing a split bus protection function of a protection relay, And said combination of said tasks not processed by said particular operator comprises a package bus protection task for the same bus and a split bus protection task for the same bus. The method of claim 6, The plurality of tasks includes a power transmission line task for realizing a power line protection relay function of each power transmission line composed of parallel lines, And said combination of said tasks not processed by said particular calculator comprises a plurality of transmission line tasks for protecting transmission lines belonging to the same parallel line. A highly reliable real-time distributed processing system for executing a plurality of tasks with a plurality of calculators connected by one or a plurality of networks, For each of the plurality of tasks, when a calculation process is executed in which a constraint or condition for completing the execution of the task is completed within a predetermined time is executed, the constraints or correlations among the plurality of arbitrary tasks or the respective constraints or correlations are different. A process of determining the degree of satisfaction of a constraint or condition to be executed by an operator is performed, whereby maximum or sufficient reliability is obtained within a range executable by the entire calculator. The method of claim 11, For each of the plurality of tasks, an execution time evaluation index for completing the execution within the predetermined time is provided, and an execution operator evaluation index for executing each of the correlations between the plurality of tasks with a different operator is provided. , The system includes an execution operator control mechanism that executes a selection process of a task to be executed and a selection process of a calculator that calculates the selected task such that the evaluation index group is maximum within a range executable by the entire calculator. High reliability real-time distributed processing system, characterized in that. The method of claim 12, The system stores in advance in the database the execution operator evaluation indicators for each of the correlations between the execution time evaluation indicators for each of the tasks and tasks other than the task, and references the indicators to perform the task selection process. And the execution operator control mechanism that executes the operator selection process. The method of claim 12, The execution operator evaluation indicators for each of the correlations between the execution time evaluation indicators for the respective tasks and the tasks other than the task are set in advance in a program module for realizing the task, and the system is configured from the task. And said execution operator control mechanism for executing said task selection process and said operator selection process, with reference to said indicator to be transmitted. The method of claim 12, The execution of each of the correlations between a symbol describing the execution time evaluation index for the task and a task other than the task, in a symbol indicating the task name, when a name for controlling each task is provided. By including a symbol describing an operator evaluation index, the system includes the execution operator control mechanism for extracting the execution time evaluation index and the execution operator evaluation index from the task name given in advance. Real time distributed processing system. The method according to any one of claims 12 to 15, The processing of the execution operator control mechanism is performed by a plurality of operators on the network. The method according to any one of claims 12 to 15, The processing of the execution operator control mechanism is performed by one or a plurality of calculators not always connected to the network. The method according to any one of claims 12 to 16, The execution operator control mechanism always monitors the operation state of the calculator, and if a change is detected in the state of any of the calculators on the network, a highly reliable real time characterized in that the task selection process and the operator selection process are executed again. Distributed processing system. The method of claim 18, The execution operator control mechanism manages the existence of a task to be generated in the future from the current calculation section and the evaluation index group related to the task, thereby making it possible to optimize the task selection process and the operator selection process through a plurality of time sections. Obtaining a high reliability real-time distributed processing system, characterized in that for performing the task scheduling. The method of claim 18, When a start request of a new task that is not included in the content controlled by the execution operator control mechanism occurs or a start instruction of a new task is issued from the outside, the execution operator control mechanism performs the task selection process and the operator selection process. High reliability real-time distributed processing system, characterized in that the re-execution. The method of claim 18, When an instruction is made for the utilization state of the calculator through an external calculator, the execution operator control mechanism assumes the utilization state of the calculator according to the contents of the instruction, thereby re-executing the task selection process and the operator selection process. High reliability real-time distributed processing system, characterized in that. The method of claim 11, An execution time evaluation index for completing the execution of each of the plurality of tasks within the predetermined time is provided, and an execution operator evaluation index for executing each of the correlations between the plurality of tasks with a different calculator is provided. The system is capable of performing an initial design that executes the selection process of the task to be executed and the selection process of the calculator that calculates the selected task such that the evaluation index group is maximum within the executable range by the entire calculator. A highly reliable real-time distributed processing system comprising a development tool. The method according to any one of claims 11 to 22, The system includes a set of communication channel monitoring tasks for verifying the normality of communication contents as the task, for each of the plurality of communication paths connecting between a plurality of computing devices, The system grasps the normality of each communication path by performing a process for determining the degree of satisfaction of a constraint or condition for executing each task composed of a set of communication channel monitoring tasks verifying the same communication path with a different calculator. High reliability real-time distributed processing system, characterized in that. The method according to any one of claims 11 to 22, The system includes a plurality of operator monitoring tasks for executing a test for checking a predetermined expression as the task and verifying a result through communication with a plurality of other calculators connected by a network, The system is capable of specifying a faulty operator or detecting a failure of the network by performing a process of determining the degree of satisfaction of a constraint or condition for executing each of the operator monitoring tasks with a different operator. Real time distributed processing system.