JPH09214532A - Shared data calculation method - Google Patents

Abstract

(57) Abstract: When performing distributed processing in a plurality of stations using shared detection data, each station executes an operation using the detection data at the same time. SOLUTION: An identifier having a logical value of 1 is registered in advance in a registration area of a memory 7 in a station having the longest arithmetic processing time, and in each station ST _{1 to} ST ₃ ,
When the arithmetic processing is completed, the registration area is referred to and it is determined whether or not the own station has the longest arithmetic processing time. As a result of the determination, the station ST ₂ having the longest processing time is the other stations ST ₁ and ST _3.
To the start command to its own station, and based on this information or command, each station
The detection data stored in the primary reception buffer 3 are transferred to the secondary reception buffer 4 all at once, and the arithmetic processing using the same detection data is executed in each station.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention, when a plurality of stations installed on a transmission path of a local area network share input data, the input data transmitted at the same time is used to perform arithmetic processing at each station. The present invention relates to an arithmetic processing method of shared data that manages the start timing of arithmetic processing in each station so as to be executed.

[0002]

2. Description of the Related Art Conventionally, in a processing station installed on a transmission path of a local area network, when a processing program shares input data, for example, multitask processing is performed to execute the processing program. An example of this will be described based on the power generation amount monitoring control device of the power plant shown in FIG. A calculation station ST ₀ , a transmission station ST ₄ and a processing station ST ₅ are installed in the transmission line 1, and the detection stations S _{a to} S _c and the driving device are respectively provided in the transmission station ST ₄ and the processing station ST _5. D _{a to} D _c are connected.
The detection apparatus S _a to S in _c example generator output, detected steam pressure, steam temperature, respectively, further, a driving device D _a to D _c in example governor valve B _a, the water supply valve B _b, the fuel valve B _c Each control valve is controlled to open and close to control the amount of power generation.

In this power generation amount supervisory control device, each detection data detected by the detection devices S _{a to} S _c is transmitted from the transmission station ST ₄ and received by the calculation station ST ₀ via the transmission line 1. , Calculation station S
At T ₀ , each detection data detected at the same time is shared and used by the multitasking processing program, the deviation between each detection data and each output target value is calculated, and further, based on each calculated deviation. The difference signal and the sum signal are obtained, and the control data in the direction in which the deviation is zero is calculated. Then, by transmitting the calculated control data through the processing station ST ₅ to the drive device D _a to D _c, based on the control data in the drive device D _a to D _c of the control valve B _a .about.B _c Opening / closing control is performed to appropriately control the rotational speed of the turbine and the generated pressure of the boiler to control the power generation amount.

As described above, the input data detected at the same time is shared by each processing program in the multitask processing program, the comparison processing is performed, and each data is calculated based on each operation result at that time. When performing the opening / closing control of the control valves B _{a to} B _c , it is necessary to be careful not to update the input data until the predetermined arithmetic processing in each processing program is completed. It is done after the low processing is completed.

[0005]

However, in the above-mentioned conventional example, since the input data cannot be updated until the arithmetic processing is completed in all the processing programs,
The calculation station has a problem that the latest data cannot be read quickly. In order to speed up the calculation process of input data and to update the input data promptly, all the processes using the input data are not performed in one station, but are distributed and performed in multiple stations. However, in this case, it is necessary to manage the read timing of the input data of each calculation station so that the input data detected at the same time can be read and processed by each calculation station. There is an unsolved problem.

Therefore, the present invention has been made by paying attention to the unsolved problem of the above-mentioned conventional example, and when performing arithmetic processing by decentralizing at a plurality of stations, the same shared input data is used. It is an object of the present invention to provide an arithmetic processing method of shared data that manages the start timing of arithmetic processing in each station so that each station can read and perform arithmetic processing.

[0007]

In order to achieve the above object, a shared data arithmetic processing method according to a first aspect of the present invention is a transmission station for transmitting data that changes sequentially to a transmission path of a local area network, A plurality of calculation stations that perform predetermined calculation processing based on the transmission data of the transmission station are connected, and each calculation station performs the calculation processing by sharing the transmission data of the transmission station. In the shared data arithmetic processing method, one station among the arithmetic stations manages arithmetic processing in each arithmetic station so as to start based on the transmission data transmitted at the same time. And

In the invention according to claim 1, one station among the plurality of arithmetic stations manages the start timing of the arithmetic processing executed in each arithmetic station, and management does not overlap. Therefore, the arithmetic processing using the transmission data transmitted at the same time is surely executed. According to a second aspect of the present invention, in the shared data arithmetic processing method, a transmission station, a plurality of arithmetic stations and a processing station are connected to a transmission path of a local area network, and each of the arithmetic stations is the transmission station. In the shared data arithmetic processing method in which the transmission data transmitted from the computer is shared and arithmetic processing is performed, and the arithmetic result is transmitted to the processing station, each of the arithmetic stations transmits the transmission data. A receiving device that is connected to a transmission line and receives transmission data that is transmitted to itself through the transmission line; and a transmission to the processing station by performing predetermined arithmetic processing based on at least the transmission data received by the receiving device. An arithmetic processing unit that calculates data and transmission data output from the arithmetic processing unit The transmitting station for transmitting to the transmission line, and a storage device in which an identifier representing a station having the longest arithmetic processing time is pre-registered, and the arithmetic station serving as a management station in which the identifier is registered in the storage device. At the end of its own processing,
Along with starting the arithmetic processing based on the transmission data, an arithmetic start command for instructing another arithmetic station to start arithmetic processing based on the same transmission data is transmitted via the transmitting device, and the arithmetic start command is issued. Another reception station for reception starts the calculation processing based on the same transmission data.

According to the second aspect of the present invention, the identifier representing the station having the longest arithmetic processing time is registered in the storage device in advance, and the arithmetic station whose identifier is registered in the storage device becomes the management station. It manages the start timing of arithmetic processing. According to a third aspect of the present invention, in a shared data arithmetic processing method, a transmission station, a plurality of arithmetic stations and a processing station are connected to a transmission path of a local area network, and each of the arithmetic stations is the transmission station. In the shared data arithmetic processing method in which the transmission data transmitted from the computer is shared and arithmetic processing is performed, and the arithmetic result is transmitted to the processing station, each of the arithmetic stations transmits the transmission data. A receiving device that is connected to a transmission line and receives transmission data that is transmitted to itself through the transmission line; and a transmission to the processing station by performing predetermined arithmetic processing based on at least the transmission data received by the receiving device. An arithmetic processing unit that calculates data and transmission data output from the arithmetic processing unit A management station is selected from among the respective calculation stations, and the remaining calculation stations send a calculation end notification to the management station when their calculation processing is completed. Along with the transmission, when a calculation start command from the management station is received, calculation processing is started based on the transmission data at the same time as that of the management station, and the management station receives data from another calculation station. When it is detected that the calculation is completed in all the calculation stations including itself based on the calculation end notification, the calculation process based on the transmission data is started, and the same transmission data is transmitted to other calculation stations. A calculation start command for instructing the start of the calculation process based on is transmitted via the transmitter. And butterflies.

According to the third aspect of the present invention, when the management station determines that the arithmetic processing is completed in all the arithmetic stations including itself based on the arithmetic end notification from the other arithmetic stations, Processing for managing the start timing of the arithmetic processing to be executed next is performed. On the other hand, when it is determined that the arithmetic processing has not ended, the processing waits until the arithmetic processing ends in all the arithmetic stations.

Further, in the shared data arithmetic processing method according to a fourth aspect of the present invention, as the management station, an arithmetic station having a maximum address or a minimum address of the addresses assigned to each arithmetic station is selected. Is characterized by. In the invention according to claim 4, since the maximum address station or the minimum address station is selected as the management station, the selection work can be easily performed.

According to a fifth aspect of the present invention, in the shared data arithmetic processing method, at least one of the arithmetic stations other than the management station is an abnormality detecting timer means that is activated at the time of starting arithmetic processing in the arithmetic processing device. And a supervisory control means for sending the operation start command when the count value of the abnormality detecting timer means exceeds a preset value longer than the longest arithmetic processing time of each arithmetic station. Characterized by being selected as a proxy station.

In the invention according to claim 5, at least one of the arithmetic stations other than the management station is selected as a management proxy station, and the management proxy station is provided with an abnormality detection timer means and a monitoring control means. Has been. Then, the management proxy station manages the start timing of the arithmetic processing when the count value of the timer means becomes larger than the arithmetic processing time of the station having the longest arithmetic processing time. Therefore, even if, for example, the maximum address station or the station having the longest arithmetic processing time that has managed the arithmetic processing fails, the management agent station can reliably manage the arithmetic processing.

According to a sixth aspect of the present invention, in the shared data arithmetic processing method, the management surrogate station has an arithmetic station having a maximum address or a minimum address among the addresses assigned to the arithmetic stations other than the management station. Is selected. In the invention according to claim 6, when, for example, the maximum address station or the minimum address station that manages the arithmetic processing fails, the maximum address station or the minimum address station among the arithmetic stations excluding the failed station Since the address station newly manages the arithmetic processing, it is possible to easily select the station for continuously executing the arithmetic processing management.

Further, in the shared data arithmetic processing method according to a seventh aspect of the present invention, the receiving device of the arithmetic station has a first buffer memory for storing transmission data from the transmitting station, and the first buffer memory for starting the arithmetic processing. A second buffer memory for transferring and holding the transmission data stored in the first buffer memory.

In the invention according to claim 7, the receiving device is provided with a receiving buffer memory which is separated into a first buffer memory and a second buffer memory, and the transmission data stored in the second buffer memory is subjected to arithmetic processing. Since it is updated only at the start, it is possible to reliably execute the arithmetic processing using the constant transmission data. The shared data arithmetic processing method according to claim 8 is such that a transmission station, a plurality of arithmetic stations and a processing station are connected to a transmission path of a local area network, and each of the arithmetic stations is the transmission station. In the arithmetic processing method of shared data in which the transmission data transmitted from the device is shared to perform arithmetic processing and the arithmetic result is transmitted to the processing station, the local area network is transmitted by the token passing transmission method. Each of the computing stations is constructed based on a receiving device that is connected to the transmission line and receives transmission data transmitted to itself through the transmission line, and at least the transmission data received by the receiving device. And perform predetermined arithmetic processing to transmit to the processing station And a transmission device for transmitting the transmission data output from the calculation processing device to the transmission path. Each calculation station continuously exchanges tokens between the stations. And the arithmetic processing time to be executed by the arithmetic processing device is set to be less than the circulation time of the token.

According to the invention of claim 8, tokens are continuously exchanged between a plurality of operation stations installed on a transmission path of a local area network configured by the token passing transmission system. Therefore, while the tokens are being exchanged between the arithmetic stations, the stations other than the arithmetic stations cannot receive the tokens and cannot transmit the data. The data is not updated. Further, in each arithmetic station, since the arithmetic processing is performed within one turn of the token, the arithmetic processing using the same transmission data can be executed.

[0018]

Embodiments of the present invention will be described below with reference to the drawings. The shared data arithmetic processing method according to the present invention is applicable to the field of monitoring and control of electric power, steel, water treatment, and the like, and as an example, a case of applying the present invention to the field of power generation monitoring and control of a power plant will be described. .

Of shared data according to the first embodiment of the present invention.
As shown in Fig. 1, the calculation processing method is based on the local area network.
Network is constructed by bus type token passing transmission method
Then, on the transmission line 1 of this network, for example, calculation of 3 stations
Station ST₁~ ST _ThreeAnd the sending station
ST_FourAnd processing station ST_FiveAre installed and sent
Credit station ST_FourThe detector S_a~ S_cIs connected
The detection device S_a~ S_cFor example generator output, steam
Detection data of pressure and steam temperature are input and processed.
Station ST_FiveDrive device D_a~ D_cIs connected
Drive device D_a~ D_cFor example, governor valve B_a,water supply
Valve B_b, Fuel valve B_cPower generation by controlling the opening and closing of each
Configured to do.

Then, each calculation station ST ₁
The ~ST _3, the receiving unit 2 for receiving detection data transmitted from the transmitting station ST ₄ is connected to the transmission line 1
And a primary reception buffer memory 3 for storing the detection data received by the receiving unit 2 while sequentially updating it, and a secondary reception buffer memory 4 for receiving the detection data from the primary reception buffer memory 3 at a predetermined timing. And an arithmetic processing unit 5 that performs arithmetic processing based on the detection data from the secondary reception buffer memory 4 and confirms the content of information received by the receiving unit 2, and transfers based on a transfer command from the arithmetic processing unit 5. A transfer processing unit 6 that generates a pulse and transfers the detection data from the primary reception buffer memory 3 to the secondary reception buffer memory 4, and a processing program that is connected to the arithmetic processing device 5 and performs an arithmetic processing of the detection data. Control data and control commands are supplied from the memory 7 as a storage device to be stored and the arithmetic processing device 5, and according to the control commands, a frame of a predetermined communication protocol is supplied. Forming a beam, and transmission device 8 for transmitting is provided with control data and a predetermined instruction information or token (transmission right). A receiving device is composed of the receiving unit 2, the primary receiving buffer memory 3, and the secondary receiving buffer memory 4.

Each calculation station ST _{1 to} ST ₃
The detection programs are shared by the respective processing programs provided in, and the respective processing stations ST _{1 to} ST ₃ execute distributed processing using the shared detection data. When performing this distributed processing, one station among the arithmetic stations ST _{1 to} ST ₃ is selected as the management station so that the distributed processing can be performed at each station using the detection data detected at the same time. This manages the start timing of the arithmetic processing, and the other stations start the arithmetic processing based on the instruction information of the arithmetic start transmitted from the management station.

The management station is selected as follows. That is, a predetermined registration area is secured in advance in each memory 7 of the arithmetic stations ST _{1 to} ST ₃ , and an identifier indicating that the station has the longest arithmetic processing time is registered in this registration area. It Then, a logical value 1 is set as an identifier in the registration area of the memory 7 of the arithmetic station having the longest arithmetic processing time at the time of initial setting, and a logical value 0 is set in the registration areas of the memory 7 of the other arithmetic stations. Is set. And
The station having the identifier of logical value 1 becomes the management station. At this time, the logical value 1 of the identifier is set by the user himself or when the created processing program is distributed to each of the operation stations ST _{1 to} ST ₃ , for example, the number of steps of the program is increased or decreased. Set automatically based on.

Next, the calculation stations ST _{1 to} ST ₃
The calculation start control process executed by the calculation processing device 5 will be described with reference to the flowchart of FIG. The calculation start control process is executed when the token is held after the calculation process is completed in each of the stations ST _{1 to} ST _3. First, in step S 1, the identifier is read by referring to the registration area of the memory 7.

Next, in step S2, it is determined whether the identifier has a logical value of 1, and it is determined whether the own station is the station having the longest arithmetic processing time. When the logical value is 0, the arithmetic start control processing is terminated, while when the logical value is 1, it is determined that the station is the station having the longest arithmetic processing time and the next step S3.
Move to

In step S3, a control command for starting the calculation is output to the transmission device 8, and the start instruction information S for instructing the start of the calculation process is transmitted from the transmission device 8. Then, in step S4, a transfer command is sent to the transfer processing unit 6 in order to transfer the detection data stored in the primary reception buffer memory 3 to the secondary reception buffer memory.

Next, in step S5, the arithmetic processing is started and the arithmetic start control processing is ended. By the calculation start control process described above, the calculation station in which the identifier of the logical value 1 is registered transmits the start instruction information S to another calculation station which is the managed station as the calculation start management station. In the managed station, when the start instruction information S is received, the arithmetic processing unit 5 confirms the content of the information, and the arithmetic processing unit 5 sends a transfer command and transmits the secondary reception buffer memory, as in step S4. Then, the detection data is transferred to, and then the calculation process is started to start the calculation in the same manner as in step S5.

Next, the operation of the first embodiment will be described. The station addresses of the arithmetic stations ST _{1 to} ST ₃ , the transmitting station ST ₄ and the processing station ST ₅ are
Stations ST ₁ , ST ₂ , ST ₃ are 3, respectively
5 and 13 and stations ST ₄ and ST ₅ are set to 4 and 10, respectively, the token T is
Stations with smaller station addresses are passed in order to larger stations.

Now, at time t shown in the time chart of FIG.
_{At 1} , the transmitting station ST ₄ receives the token T and sets the detection data D ₁ including the generator output, the steam pressure, and the steam temperature into a predetermined frame, and each calculation station S.
When transmitting to T _{1 to} ST ₃ , each station ST ₁ to ST ₃
ST ₃ confirms the content of the received frame, determines that it is the data transmitted to its own station, and outputs the detected data D _{1 one} by one.
The data is stored in the next reception buffer memory 3.

Here, the calculation stations ST _{1 to} ST
_{In 3} , in the initial state until the station having the longest arithmetic processing time transmits the arithmetic processing start instruction information S, the detection data D ₁ is automatically received in the primary reception buffer memory 3 in the initial state. It is set so that the data is transferred to the secondary reception buffer memory 4 and the arithmetic processing is started.
As a result, immediately after the time t ₁ , the arithmetic processing using the desired data in the detection data D ₁ is executed in each of the arithmetic stations ST _{1 to} ST ₃ .

For example, in the calculation station ST ₁ ,
The deviation between the generator output value and the output target value is calculated, the deviation between the steam pressure value and the pressure target value is further calculated, and the governor valve control value is calculated from the difference between these deviations. It calculates a control data D _1a for opening and closing control of the governor valve B _a from each data generator output and the steam pressure in the D _1. In the calculation stations ST ₂ and ST ₃ , the data of the generator output, the steam pressure, and the steam temperature in the detection data D ₁ are used to perform different calculations, and the water supply valve B
_b and the control data D _1b and D for controlling the opening and closing of the fuel valve B _c
Calculate each _1c .

The calculation stations ST _{1 to} ST _{3 are also provided.}
In the length of processing time to be executed is different at each station, as shown in FIG. 3, a minimum processing time at station ST _3, the station ST _1, the station ST ₂ and the processing time is gradually lengthened Go. During this calculation processing time, the detection data D ₂ and D ₃ are sequentially transmitted from the transmitting station ST ₄ , the detection data in the primary reception buffer memory 3 are sequentially updated, and the latest detection data is stored.

Then, the respective calculation stations ST _{1 to} S
In T _3, the calculation processing is completed refers to the registration area of the memory 7, determines whether the identifier of the logical value 1 is registered. Here, since the registration area of the calculation station ST ₂ is the identifier of the logical value 1 is set, this station ST ₂ is at time t ₂ when holding the first token after processing ends, the station ST ₁ And ST ₃
To the start station, and also sends an instruction to start the arithmetic processing to its own station. Station ST ₁
In ST ₃ and ST ₃ , the start instruction information S is received and the detection data of the primary reception buffer memory 3 is transferred to the secondary reception buffer memory 4. As a result, in each of the arithmetic stations ST _{1 to} ST ₃ , arithmetic processing using desired data in the same detection data D ₄ is executed.

Then, the previous calculation result using the same detection data D ₁ is transmitted as control data by multitask processing in each calculation station when the token is held after the start instruction information S is transmitted. . That is, at time t ₃ , the calculation result using the detection data D ₁ is transmitted from the station ST ₃ as control data D _1c and is received by the processing station ST ₅ to control the opening / closing of the fuel valve B _c . Similarly, at the time t ₄ , the calculation station ST
_The control data D _1a (calculation result of the detection data D ₁ ) transmitted from ₁ is received by the processing station ST ₅ and the governor valve B _a is controlled to open and close. Further, at time t ₅ , the control data D _1b (calculation result of the detection data D ₁ ) transmitted from the station ST ₂ is received by the processing station ST ₅ and the water supply valve B _b is controlled to open and close. Then, the detection data in the primary reception buffer memory 3 is sequentially updated, and when the next start instruction information S is transmitted at time t ₆ , the arithmetic processing using the desired data in the detection data D ₇ is executed. To be done.

As described above, according to the first embodiment, the station ST ₂ having the longest arithmetic processing time transmits the start instruction information S and outputs the arithmetic processing start command to its own station, Since the detection data of the next reception buffer memory 3 is transferred to the second reception buffer memory 4, when the start instruction information or the like is output, the stations ST ₁ and S
At T ₃ , the arithmetic processing has already been completed, and it becomes possible to execute the arithmetic processing using the same detection data without adversely affecting the arithmetic processing.

Next, a shared data arithmetic processing method according to the second embodiment of the present invention will be described. In the second embodiment, the arithmetic processing is managed by the station having the maximum address among the arithmetic stations ST _{1 to} ST ₃ . The connection of each component of the second embodiment is the same as that of FIG. 1, but in each memory 7 of the arithmetic stations ST _{1 to} ST ₃ , instead of the identifier registration area, an address table as shown in FIG. 4 is used. Are provided respectively, and the second embodiment is configured similarly to the first embodiment except this point.

Addresses of all stations that perform arithmetic processing are stored in advance in the address table when the network is constructed, and an address table is created. That is, an addressable area is secured in a predetermined storage area of the memory 7, a bit number is determined in order for each bit in this area, and a logical value 1 is set in the area of the bit number corresponding to the address value of each station. The address table is set up.

In this case, since the address value of the arithmetic station ST ₁ is 3, the logical value 1 is set in the third bit, and similarly, the logical value 1 is set in the fifth bit and the 13th bit. Is set. The logical value at this time is set by the user, or is automatically set when the created processing program is distributed to the arithmetic stations ST _{1 to} ST ₃ .

Next, the processing at the time of transmission and the processing at the time of reception constructed by the arithmetic processing unit of each arithmetic station will be described. The processing program for transmission stored in the memory 7 is the same in each of the arithmetic stations ST _{1 to} ST ₃ , but the processing actually executed differs between the maximum address station and other stations. At the time of system start-up, as a pre-stage of the process at the time of transmission, the discrimination process for discriminating whether or not it is the maximum address station shown in the flowchart of FIG. Determine if it is an address station. For example, if the logical value 1 is not set in the area of the bit number larger than the address value of the own station by referring to the address table, the own station is determined to be the maximum address station.

When the own station is the maximum address station, the step
Control flag F ₁Set logical value 1 to
On the other hand, if it is not the maximum address station, step S
13, the control flag F₁Set logical value 0 to
Then, the determination process is ended. Next, after finishing the arithmetic processing
When the token is held, the transmission shown in the flowchart of FIG. 6 is performed.
Processing is performed. First, in step S21, the control flag is set.
Lug F₁And read F₁Determine if = 1
You. F₁= 0, stations that are not maximum address stations
To execute processing as an option to step S22.
Transition.

In step S22, an end notification command is output to the transmitter 8 to cause the transmitter 8 to transmit the operation end notification information to another arithmetic station. Then, the process shifts to step S23, the logical value 1 is written in the area of the same bit number as the address of the own station in the end notification table of the own station, and the transmission process as the station which is not the maximum address station is completed. Here, the end notification table is created in a predetermined storage area of the memory 7 of each of the calculation stations ST _{1 to} ST _{3 in} the same format as the address table shown in FIG. When the station finishes the computation or receives the computation end notification information as described later, the logical value 1 is written in the area of the bit number corresponding to the address of the station.

On the other hand, when F ₁ = 1, the process shifts to step S24 to execute the process as the maximum address station, and as the process at the end of the calculation, the same bit number as the address value of the own station is displayed in the end notification table. Logical value 1 in area
Is written, and the process proceeds to step S25. Step S25
Then, the end notification table in which the station for which the arithmetic processing is completed is stored is referred to.

[0042] Then, the process proceeds to step S26, it is determined whether processing has been completed in operation for the station ST ₁ ~ST ₃ based on the logical value written to completion notification table. Here, for example, the respective bits of the address table and the end notification table are compared, and when the logical values of the respective bits of both tables are the same value, the arithmetic processing is performed in each of the arithmetic stations ST _{1 to} ST _3. When it is determined that the process is completed, the process proceeds to the next step S27. On the other hand, when the logical values of the bits of both tables are not the same,
The process ends.

In step S27, a control command for starting the calculation is output to the transmission device 8 and the start instruction information S for instructing the start of the calculation process is transmitted from the transmission device 8. Then, in step S28, a transfer command is output to the transfer processing unit 6 in order to transfer the detection data stored in the primary reception buffer memory 3 to the secondary reception buffer memory 4.

Next, in step S29, the end notification table of the own station is cleared and the logical value 0 is set in the area of all bit numbers. Next, the process proceeds to step S30, the arithmetic processing is started, the arithmetic using the detection data transferred to the secondary reception buffer memory 4 is started, and the processing at the time of transmission is ended.

Next, the processing at the time of reception will be described based on the flowchart shown in FIG. The process at the time of this reception is started by the upper main program when the data is received at the arithmetic station.
Then, it is determined whether or not the calculation end notification information is received. When the calculation end notification information is received, the process proceeds to step S32, and the logical value 1 is written in the area of the bit number corresponding to the address value of the own station in the end notification table of the own station,
Return to the upper main program.

On the other hand, when it is determined that the received data is not the calculation end notification information, the process proceeds to step S33,
It is determined whether the start instruction information S has been received. When it is determined that the start instruction information S has not been received, the process returns to the upper main program. On the other hand, when it is determined that the start instruction information S has been received, the process proceeds to the next step S34.

In step S34, the above-mentioned step S2 is performed.
Similarly to 8, the transfer command is output. And step S
In step S35, the end notification table of the own station is cleared as in step S29. Next, step S36.
Then, the arithmetic processing is started in the same manner as in step S30 described above, and the process returns to the upper main program. Next, the operation of the second embodiment will be described.

Time t shown in FIG.₁Then, the station for sending
ST_FourReceived tokens, generator output, steam pressure,
Detection data D consisting of steam temperature₁Station for each calculation
ST₁~ ST_ThreeTo each station ST₁
~ ST_ThreeThen, the detection data D ₁Use the desired data in
Similar to the first embodiment, the previously described arithmetic processing is executed.
You.

After the arithmetic processing is started, the station ST
₃ first ends the arithmetic processing, and the station ST ₃ writes the logical value 1 in the area of the bit number corresponding to the address of the own station in the end notification table of the own station. next,
At time t ₇ , the station ST ₁ retains the token after completing the arithmetic processing, transmits the arithmetic end notification information E _ST1 to the stations ST ₂ and ST ₃ , and, at the same time as described above, the corresponding bit in the end notification table. Write a logical value of 1 in the number area. The stations ST ₂ and ST ₃ receive the operation end notification information E _ST1 and write the logical value 1 in the area of the corresponding bit number in the end notification table.

Then, at time t ₈ , the station ST ₂
When the token is held after the arithmetic processing is completed, the arithmetic end notification information E _ST2 is transmitted to the stations ST ₁ and ST ₃ and, in the same manner as described above, the logical information is stored in the area of the corresponding bit number of the own end notification table. Set the value to 1. Upon reception, the stations ST ₁ and ST ₃ write the logical value 1 in the area of the corresponding bit number in the end notification table in the same manner as above.

Next, at time t ₉ , the station ST ₃ of the maximum address station judges whether or not it has received the operation end notification information from all the other stations when holding the token. At this time, since the calculation end notification information E _ST1 and E _ST2 have been received, the station ST ₃ is
₁ , the calculation processing start instruction information S is transmitted to ST ₂ and the calculation processing of the own station is started. Station ST ₁ ,
In ST ₂ , the start instruction information S is received and the arithmetic processing is started. As a result, all calculation stations ST
_{In 1 to} ST ₃ , arithmetic processing using desired data in the same detection data D ₁ is executed.

Further, from the stations ST _{1 to} ST ₃ , the control data D _1a , D based on the previous detection data D ₁ is held when the token is held after the start instruction information S is transmitted as in the first embodiment. _1b and D _1c are sequentially transmitted, and the opening / closing control of each control valve B _a , B _b , B _c is executed. Then, at the start of the arithmetic processing, the end notification table of each of the stations ST _{1 to} ST ₃ is cleared, and the detection data in the primary reception buffer memory 3 are sequentially updated so that the next start instruction information S at time t _10. When sent, detect data D
Arithmetic processing using desired data in ₇ is executed.

As described above, according to the second embodiment, the maximum address station transmits the start instruction information S after confirming that the arithmetic processing is completed at each of the arithmetic stations ST _{1 to} ST _3. Since the arithmetic processing start command is output to the local station and the detection data of the primary reception buffer memory 3 is transferred to the secondary reception buffer memory 4, the arithmetic processing is started without adversely affecting the arithmetic processing. Timing can be managed, and each operation station ST _{1 to} S
At T ₃ , it becomes possible to execute the arithmetic processing using the same detection data.

Further, the maximum address station is selected as the management station for transmitting the start instruction information S, and the maximum address station can be easily extracted by referring to the address table. In the second embodiment, the transmission of the start instruction information S is performed by the maximum address station, but the present invention is not limited to this, and may be performed by the minimum address station. Also in this case, the minimum address station can be easily extracted by referring to the address table. In addition to the above, any one station may be selected as the station that transmits the start instruction information S.

In the second embodiment, the calculation end notification information is transmitted at the calculation station excluding the maximum address station, but the maximum address station may also transmit the calculation end notification information. . Next, a shared data arithmetic processing method according to the third embodiment of the present invention will be described. In the third embodiment, when the maximum address station which manages the arithmetic processing cannot execute the arithmetic processing management due to the occurrence of an abnormality due to a temporary runaway of the arithmetic processing, the arithmetic operation is performed by the alternative arithmetic station. It manages the processing. As shown in the block diagram of FIG. 9, a counter circuit 9 serving as a watchdog timer as abnormality detecting timer means is connected to the arithmetic processing units 5 of the arithmetic stations ST _{1 to} ST ₃ , and this counter circuit 9
An abnormality is detected based on the count value of 1 and the processing for dealing with the abnormality is performed. Except for this point, each component is configured similarly to the second embodiment.

The counter circuit 9 is reset to zero and starts counting each time the arithmetic processing is started by the reset signal from the arithmetic processing unit 5. Then, for example, the pulses generated by the crystal oscillator are counted, and the count value is output to the arithmetic processing device 5. Further, from the counter circuit, a signal for forcibly stopping and resetting the arithmetic processing when the time is up is supplied to the arithmetic processing device 5, and the maximum address station which is the management station has runaway and timed up. At times, it is reset and a new arithmetic process is executed from the initial state.

Next, the monitoring control processing as the monitoring control means executed by the arithmetic processing unit will be described with reference to the flowchart of FIG. This monitoring control process is subsequently executed after the transmission process shown in FIG. 6 of the second embodiment is executed when the token is held. The processing program for this supervisory control processing is built in each arithmetic station, but since the actual processing is executed by stations other than the maximum address station, first in step S41, F ₁
It is determined whether or not = 0. When F ₁ = 1, the own station is the maximum address station, so the processing ends, while
When F ₁ = 0, the process proceeds to step S42.

In step S42, the count value of the counter circuit 9 is read. Next, in step S43, it is determined whether or not the count value has timed up. Here, the count value and the set value are compared, and the set value is set by adding the time considering the delay time in the transmission protocol to the operation processing time of the operation station having the longest operation processing time. To be done. Then, if the count value is smaller than the set value, it is determined that the time has not expired and the monitoring control process is terminated. On the other hand, if the count value is greater than the set value, it is determined that the time has expired. The process moves to S44.

In step S44, the address table is referenced. Next, in step S45, it is determined whether or not the own station is the station having the second largest address among the stations for calculation. By searching for the bit in which the logical value 1 is described in the address table and comparing the second largest bit number among the searched bits with the address value of the own station, the own station is normally operated at the present time. It can be determined whether it is the maximum address station that is operating.
When the address value of the own station becomes the same as the second largest bit number and it is determined that the own station is the maximum address station in the normal operating state, the process proceeds to the next step S46. On the other hand, when it is not the maximum address station in the normal operating state, the monitoring control process is ended.

In step S46, the process of starting the calculation is executed, and the same processes as those in steps S24 to S30 are performed. That is, the logical value 1 is written in the corresponding bit number of the end notification table, the end notification table is referred to, it is determined whether or not the operation is completed in all stations except the maximum address station, and the operation is completed. If it is determined that the transmitter 8
A control command to start the calculation is output to, the start instruction information S of the calculation process is transmitted to another calculation station, and the start instruction is also given to the own station. Then, the clear processing of the end notification table is performed, and further the arithmetic processing is started. Then, the monitoring control process is ended.

Next, the operation of the third embodiment will be described. Normally, the arithmetic station rarely fails, so the maximum address station is the same as in the second embodiment.
The processing of the calculation start instruction shown in steps S24 to S30 is executed to manage the calculation processing. On the other hand, stations other than the maximum address station execute the supervisory control processing shown in FIG. 10 when holding the token. Since the counter circuit 9 is not reset when the maximum address station temporarily runs away, for example, at time t ₁₁ shown in the time chart of FIG. 11, the count value of the counter circuit 9 becomes larger than the set value and reset to the arithmetic processing unit of the station ST _3. A signal is supplied, and the arithmetic processing of the station ST _{3 which has} runaway is forcibly stopped. Then, in the arithmetic stations ST ₁ and ST ₂ , for example, when the token is held, the address table is referred to and it is determined whether or not the own station is the maximum address station among the stations in the normal operating state.

Here, the calculation station ST ₂ is selected as the maximum address station, the station ST ₂ refers to the end notification table and confirms that the calculation is completed in the stations ST ₁ and ST ₂ , and at time t ₁₂ . , The station ST ₁ and ST ₃ are transmitted with the instruction S to start the arithmetic processing, and the station itself is instructed to start the arithmetic processing. Then, the control data is not output at the station ST _{3 because} the arithmetic processing is stopped midway at the station ST _3.
And from ST ₂ control data D based on the detected data D _n
na and D _nb are output respectively. After this, since the counter circuit 9 does not time up due to the runaway being stopped, the station ST ₃ which is the original maximum address station is
Becomes the management station again and transmits the start instruction information S for the arithmetic processing.

As described above, according to the third embodiment, when the maximum address station that manages the arithmetic processing fails and is in a runaway state, for example, when the counter circuit 9 times out, the normal operation is performed. Since the maximum address station among the stations in the state is in charge of managing the arithmetic processing, it is possible to surely execute the arithmetic processing using the same detection data. When the counter circuit times out, the arithmetic processing is not managed, but the maximum address station in the normal operating state is selected before the arithmetic processing is managed. This has the advantage that the management can be performed and the analysis when a failure occurs becomes easy.

In the third embodiment, the monitoring control process is performed by the maximum address station among the stations in the normal operating state, but it may be performed by the minimum address station.
Further, it may be performed in one station selected based on a predetermined rule, and in these cases, the above effect can be obtained. Further, in the third embodiment, the maximum address station manages the arithmetic processing as in the second embodiment until the count value exceeds the set value, but the present invention is not limited to this. As in the first embodiment, the station having the longest arithmetic processing time may manage the arithmetic processing. In this case, since the address table is not created, after the count value exceeds the set value, for example, the station that first holds the token sends the start instruction information S to each station to manage the arithmetic processing. Try to run.

In the third embodiment, the counter circuit is used as the abnormality detecting timer means to count the arithmetic processing time, but the invention is not limited to this.
For example, a software timer that increments the variable value by one for each predetermined timer interrupt process may be configured to count the arithmetic processing time. Next, a shared data arithmetic processing method according to the fourth embodiment of the present invention will be described.

In the fourth embodiment, as shown in FIG. 12, for example, a bus-type token passing transmission system is used so that tokens are continuously exchanged between the arithmetic stations ST _{1 to} ST _3. Are continuously allocated. Then, each component is configured similarly to the first embodiment. Here, the calculation station S
Address of T ₁ ~ST ₃ is set to each of 3, 4, 5, the transmitting station ST _4, the address of the processing station ST _5, are each set to 10 and 13.

For example, now, the transmitting station ST ₄
There receives the token and transmits the detection data D ₁ in each operation for stations ST ₁ ～ST _3, to the primary reception buffer memory 3 of calculating the station ST ₁ ～ST ₃ is detected data D ₁ is stored . Then, when the token is transferred from the transmitting station ST ₄ to the processing station ST ₅ , the processing station ST ₅ immediately transfers the token to the arithmetic station ST ₁ because there is no data to be transmitted.

When receiving the token, the arithmetic station ST ₁ transfers the detection data D ₁ of the primary reception buffer memory 3 to the secondary reception buffer memory 4 in synchronization with the reception of the token, and starts the arithmetic processing. To do. Similarly,
Upon receiving the token, the arithmetic stations ST ₂ and ST ₃ also transfer the detection data D ₁ of the primary reception buffer memory 3 to the secondary reception buffer memory 4 and start the arithmetic processing. Then, each arithmetic processing using the detection data is set to be completed while the token makes one round.

As described above, according to the fourth embodiment, since the station addresses are continuously allocated so that the tokens are continuously transferred between the arithmetic stations ST _{1 to} ST ₃ , the tokens are assigned to the stations. ST ₁ ~
No detection data is transmitted from the transmitting station ST ₄ when the data is transmitted and received between ST ₃ , and when the token is received, the detection data in the primary reception buffer memory 3 is transferred to the secondary reception buffer memory 4 for calculation. Since the processing is started, the arithmetic processing can be executed using the same detection data.

In the fourth embodiment, the case of the bus type token passing transmission system has been described.
It is also applicable to a ring type token passing transmission system, in which case each station is arranged so that tokens are continuously exchanged between the operation stations ST _{1 to} ST ₃ . In the above-mentioned first to fourth embodiments, although running processing using the same shared data by using the detection data supplied from the transmission station ST _4, for example, transmission path server computer 1 Alternatively, the shared data supplied from the server computer may be used to execute the arithmetic processing using the same shared data.

Further, in the first to third embodiments, the same
Management of arithmetic processing using shared data of
Transmission of the calculation start information S, the primary reception buffer
The detection data stored in the memory is used as a secondary reception buffer memo.
But I am not limited to this.
It is not something to be done. For example, the transmitting station ST _Four
Sends detection data only when there is a request to send
After the calculation is completed at all stations,
And one management station is the transmitting station ST_Four
A transmission request is output to the
Option ST_FourIs each station ST₁~ ST_ThreeDetected in
Send the data. In this case, during calculation processing
The data in the primary receive buffer memory may be updated at
Since it does not exist, the detection data of the primary reception buffer memory is used.
The same shared data can be processed by
The secondary reception buffer memory can be omitted.

In the first to third embodiments described above, the case of the bus type token passing transmission system has been described, but it is also applicable to the ring type token passing transmission system, and further, the CSMA / CD transmission system. It is also applicable to so-called Ethernet.

[0073]

As described above, according to the shared data arithmetic processing method according to the first aspect, one station among a plurality of arithmetic stations uses the same shared data in each arithmetic station. Since the management is performed so as to perform the calculation and the management is not duplicated, there is an effect that the calculation processing using the same shared data can be surely executed.

According to the shared data arithmetic processing method of the second aspect, the identifier representing the arithmetic station having the longest arithmetic processing time is registered in advance in the storage device, and the identifier is registered in the storage device. Since the arithmetic station becomes a management station and manages the start timing of arithmetic processing, there is an effect that distributed processing using the same transmission data can be realized by a plurality of arithmetic stations.

According to the shared data arithmetic processing method of the third aspect, the management station performs arithmetic operations on all arithmetic stations including itself based on the arithmetic end notification from the other arithmetic stations. When it is determined that the processing is completed, the arithmetic processing to be executed next is started in each arithmetic station, so that it is possible to easily and surely realize the distributed processing using the same shared data. Have.

Further, according to the shared data arithmetic processing method of the fourth aspect, since the maximum address station or the minimum address station is selected as the management station, the selection work can be easily performed. According to the shared data arithmetic processing method of the fifth aspect, at least one of the arithmetic stations other than the administrative station is selected as the administrative agent station, and in this administrative agent station, the abnormality detecting timer means is counted. Since the arithmetic processing is started when the numerical value becomes longer than the arithmetic processing time of the station having the longest arithmetic processing time, the arithmetic processing is managed, for example, the maximum address station or the longest arithmetic processing time is set. Even if a station fails,
This has the effect that the management agent station can reliably manage the arithmetic processing.

According to the shared data operation processing method of the sixth aspect, when the station that manages the operation processing fails, for example, the maximum of the operation stations excluding the failed station is used. Since the address station or the minimum address station newly manages the arithmetic processing, it is possible to easily extract the station for executing the management of the arithmetic processing when a failure occurs.

Further, according to the shared data arithmetic processing method of the seventh aspect, the transmission data stored in the second buffer memory is not updated during the arithmetic processing but only when the arithmetic processing is started. This has the effect that the arithmetic processing can be executed using constant transmission data. And according to the shared data arithmetic processing method according to claim 8, particularly when the local area network is constructed by the token passing transmission method, between the arithmetic stations connected to the transmission path of this network,
Since tokens are sent and received continuously and the arithmetic processing time is set to less than the token circulation time, each arithmetic station surely and easily executes arithmetic processing using the same detection data. It has the effect of being able to.

[Brief description of drawings]

FIG. 1 is a block diagram showing a configuration of an arithmetic processing system of shared data according to the present invention.

FIG. 2 is a flowchart showing management of arithmetic processing according to the first embodiment.

FIG. 3 is a time chart showing the operation of the first embodiment.

FIG. 4 is an explanatory diagram showing configurations of an address table and an end notification table.

FIG. 5 is a flowchart showing a determination process of a maximum address station according to the second embodiment.

FIG. 6 is a flowchart showing processing at the time of transmission according to the second embodiment.

FIG. 7 is a flowchart showing a process at the time of reception according to the second embodiment.

FIG. 8 is a time chart showing the operation of the second embodiment.

FIG. 9 is a block diagram showing a configuration of a third embodiment.

FIG. 10 is a flowchart showing a monitoring control process according to the third embodiment.

FIG. 11 is a time chart showing the operation of the third embodiment.

FIG. 12 is a block diagram showing a configuration of a fourth embodiment.

FIG. 13 is a block diagram showing a configuration of a conventional example.

[Explanation of symbols]

DESCRIPTION OF SYMBOLS 1 transmission line 2 receiving part 3 primary receiving buffer memory 4 secondary receiving buffer memory 5 arithmetic processing device 6 transfer processing part 7 memory 8 transmitting device 9 counter circuit ST _{1 to} ST ₃ arithmetic station ST ₄ transmitting station ST ₅ process For station

Claims (8)

[Claims] 1. A transmission station for transmitting successively changing data and a plurality of calculation stations for performing a predetermined calculation process based on the transmission data of the transmission station are connected to a transmission path of a local area network. In the shared data arithmetic processing method, wherein each arithmetic station shares the transmission data of the transmitting station to perform arithmetic processing, one station among the arithmetic stations is one of the arithmetic stations. Is managed so as to be started based on the transmission data transmitted at the same time. 2. A transmission station, a plurality of calculation stations, and a processing station are connected to a transmission path of a local area network, and the transmission data that is transmitted from the transmission station is changed at each calculation station. In the shared data arithmetic processing method in which the arithmetic processing is shared and the arithmetic result is transmitted to the processing station, each of the arithmetic stations is connected to the transmission path and is connected via the transmission path. A receiving device for receiving transmission data transmitted to itself; an arithmetic processing device for performing a predetermined arithmetic process based on at least the transmission data received by the receiving device to calculate transmission data for the processing station; A transmission device for transmitting the transmission data output from the arithmetic processing device to the transmission path, and A storage device in which an identifier representing a station having a calculation processing time is registered in advance, and the calculation station as a management station in which the identifier is registered in the storage device, when its own calculation process ends, Along with starting the arithmetic processing based on the transmission data, an arithmetic start command for instructing another arithmetic station to start arithmetic processing based on the same transmission data is transmitted via the transmitting device, and the arithmetic start command is issued. The other calculation station that has received starts calculation processing based on the same transmission data, the calculation processing method of shared data. 3. A transmission station, a plurality of calculation stations, and a processing station are connected to a transmission line of a local area network, and the transmission data transmitted from the transmission station is changed at each calculation station. In the shared data arithmetic processing method in which the arithmetic processing is shared and the arithmetic result is transmitted to the processing station, each of the arithmetic stations is connected to the transmission path and is connected via the transmission path. A receiving device for receiving transmission data transmitted to itself; an arithmetic processing device for performing a predetermined arithmetic process based on at least the transmission data received by the receiving device to calculate transmission data for the processing station; A transmission device for transmitting transmission data output from the arithmetic processing device to the transmission path A management station is selected from among the respective calculation stations, and the remaining calculation stations transmit a calculation end notification to the management station when their calculation processing is completed, and calculate from the management station. When a start command is received, arithmetic processing is started based on the transmission data at the same time as the management station, and the management station includes all of itself based on the calculation end notification from another calculation station. When it is detected that the calculation at the calculation station is finished, the calculation process based on the transmission data is started, and the calculation start is instructed to other calculation stations to start the calculation process based on the same transmission data. A shared data arithmetic processing method, characterized in that a command is transmitted via the transmission device. 4. The shared data according to claim 3, wherein the management station is selected as a calculation station having a maximum address or a minimum address of the addresses assigned to each calculation station. Arithmetic processing method. 5. At least one of the arithmetic stations other than the management station has an abnormality detection timer means started at the start of arithmetic processing in an arithmetic processing device, and a count value of the abnormality detection timer means is preset. It is selected as a management surrogate station having a supervisory control means for sending the calculation start command when the set value exceeds the longest calculation processing time of each set calculation station. Item 5. The shared data arithmetic processing method according to any one of items 2 to 4. 6. The management station is selected as a management station having a maximum address or a minimum address of the addresses assigned to the respective calculation stations other than the management station. Calculation method of shared data described. 7. The receiving device of the computing station comprises:
A first buffer memory for storing transmission data from the transmitting station and a second buffer memory for transferring and holding the transmission data stored in the first buffer memory at the start of arithmetic processing are provided. 7. The shared data arithmetic processing method according to claim 2, wherein: 8. A transmission station, a plurality of calculation stations, and a processing station are connected to a transmission path of a local area network, and the transmission data that is transmitted from the transmission station is changed at each calculation station. In a shared data arithmetic processing method in which the arithmetic processing is shared and the arithmetic result is transmitted to the processing station, the local area network is constructed by a token passing transmission method, and each of the arithmetic stations is Is a receiving device that is connected to the transmission line and receives transmission data that is transmitted to itself through the transmission line; and a predetermined arithmetic process based on at least the transmission data received by the receiving device to perform the processing. And an arithmetic processing unit for calculating transmission data for the station A transmission device for transmitting transmission data output from the processing device to the transmission path, wherein each of the arithmetic stations is set to continuously exchange tokens between the stations, and the arithmetic processing is performed. An arithmetic processing method for shared data, wherein the arithmetic processing time executed by the device is set to be less than the token circulation time.