JPH0377114A - Time coincidence system for decentralized system and train operation control system - Google Patents

Abstract

PURPOSE:To secure the autonomous performance to each element in terms of the time by selecting the system time out of the time information on all communication stations by majority and correcting the time of its own station to the system time when the deviation exceeds the set value at each communication station. CONSTITUTION:Each communication station produces the time of its own and at the same time transmits successively the information on the produced time to other communication stations 12 and 14. Then the system time serving as the reference time of a system is selected out of the time information on all stations 10, 12 and 14 by majority. Each of these stations 10 - 14 compares its own time with the system time. Then the time of its own station 10 is corrected to the system time when the deviation the time of the station 10 and the system time exceeds the set value. The station 10 produces a new time based on the corrected time. Thus it is possible to autonomously detect the time abnormality for each element and also to correct the time.

Description

[Detailed Description of the Invention] [Industrial Application Field] The present invention relates to a time synchronization method and a train operation management system for a distributed system, and particularly relates to a time synchronization method and a train operation management system for a distributed system, and in particular, to a system for controlling the time synchronization of a distributed system, and in particular for a plurality of communication stations, a plurality of computers, and a plurality of station communication stations. The present invention relates to a time synchronization method for a distributed system and a train operation management system suitable for synchronizing the time of each part with the system time in a system in which the system is distributed.

[Conventional technology]

In recent years, with the remarkable development of microcomputers, the mainstream of computer systems has become load-sharing distributed systems and autonomous distributed systems in which multiple small computers are connected via communication paths. In these distributed systems, the computers are connected to each other through talk-ring broadcast communication channels, which enables risk distribution and the construction of systems at low cost.

[Problem to be solved by the invention]

However, in conventional distributed systems, no consideration has been given to unifying the system time, which is the reference time of the system, and if this system is used as a train operation management system, there is a risk that abnormalities in the computer time at each station may not be detected. There is. In other words, in conventional systems, a standard clock is installed in a single computer among distributed computers, the system time is set according to this standard clock, and this time information is sent to each computer via a broadcast communication channel. It is now transmitted to For this reason, the computer receiving the time information cannot correct the time, and even if a time abnormality occurs in a steady state, there is a problem in that the abnormality cannot be detected and corrected. Namely.

Detection of a time abnormality will have to wait until the standard time is sent in the next cycle. In particular, in a system that performs control based on time, it is necessary to keep the processing time of each computer making up the system within the allowable time error of the system. Therefore, as described in Japanese Patent Application Laid-open No. 60-24983, as a processing method based on absolute time, a central computer communicates the relative time with respect to the control start time to other distributed computers. Therefore, a method has been proposed that synchronizes processing with respect to absolute time, but even in this method, time is centrally managed, and each computer in a distributed arrangement cannot process time information independently.

As described above, conventional distributed systems have the disadvantage that it is not possible to constantly monitor the time of all the computers that make up the system, and that it is not possible to quickly respond to time abnormalities.

Furthermore, since time is centrally managed, if a time anomaly occurs in the central computer, the time anomaly will spread to all computers making up the system.

An object of the present invention is to provide a time synchronization method and a train operation management system for a distributed system in which the elements constituting the system can individually detect and correct time abnormalities.

[Means to solve the problem]

In order to achieve the above object, the present invention provides, as a first time matching method, a distributed system in which a plurality of distributed communication stations are connected to each other via an information transmission path.
Each communication station generates its own time and sequentially transmits information about the generated time to other communication stations, and selects the system time, which will become the system's reference time, by majority vote from among the time information of all communication stations. , each communication station compares the system time and its own time, and if the deviation exceeds a set value, it corrects its own time to the system time and generates a new time based on the corrected time. This system employs a time synchronization method for distributed systems.

The second time synchronization method is a distributed system in which a plurality of distributed computers are connected to each other via an information transmission path, in which each computer generates a time and transmits information about the generated time to other computers. The system time is transmitted sequentially to the computers, and the system time that becomes the system's reference time is selected by majority vote from among the time information of all computers.The system time and the generated time are compared on each computer, and the deviation is determined from the set value. When it exceeds the time, correct the generated time to the system time,
This system employs a distributed system time matching method that generates a new time based on the corrected time.

As a train operation management system, multiple station communication stations distributed in a distributed manner are connected to each other via information communication channels, and each station communication station generates its own time and transmits information about the generated time to other station communication stations. The system time, which will be the standard time of the system, is selected by majority vote from among the time information of all the station communication stations, and each station communication station compares the system time with its own time. When the deviation exceeds the set value, the local station's time is corrected to the system time, a new time that matches the system time is generated based on the corrected time, and each station trains the train according to the local station's time and the timetable schedule. The system is designed to manage the operation of the train.

[Effect]

The time status is exchanged between the elements that make up the system, and the system time is selected by majority vote from among the time lines of all elements, and the time of each element is different from the system time, and the difference exceeds the set value. In some cases, the time of each element was corrected to the system time, and a new time was generated based on the corrected time, making it possible for each element to autonomously detect time abnormalities and correct the time. Become.

〔Example〕

Hereinafter, one embodiment of the present invention will be described based on the drawings.

In FIG. 1, a plurality of computers 10, 12, 14 are distributed along a broadcast transmission path 16, and information is exchanged between the computers using the transmission path 16 as an information transmission path. ing. Each computer 10,
12.14 is provided with a time data reception section 18, a majority decision section 20, a system time selection section 22, a time update section 24, a time generation section 26, a time check correction section 28, and a time data transmission section 30.

The time data receiving section 18 receives time data from the transmission line 16 and outputs the input time data to the majority decision section 20. The majority decision section 20 decides whether or not the time data of at least two computers match among the input time data, and outputs the decision result to the system time selection section 22 . The system time selection unit 22 is configured to select a system time that will be the reference time of the system based on the determination result from the majority decision unit 20. In other words, when the time data of two of the three computers match, the time update unit 24 selects the time according to this time data as the system time. Based on this, a time update command is output at a cycle (time update cycle) that indicates the minimum time resolution, for example, at a 3 second cycle.

The time generation unit 26 generates the time based on the output of the counter,
The generated time data is output to the time check correction section 28. The time check correction section 28 uses the time data from the time generation section 26 and the system time selection section 22.
system time data from , and determines whether the generated time matches the system time, and if the deviation exceeds the set value, the generated time is corrected to the system time. There is. The time check correction section 28 outputs data regarding the generated time or the corrected time, and this data is output from the time data transmission section 30 to the transmission line 16.

Among the various arithmetic operations performed by each computer 10, 12, 14, majority decision processing for time data is performed by a majority decision program, and specifically, is carried out according to the flowchart shown in FIG. That is, the transmission line 16
When time data is input from , the time data is read out according to the one time data reception program (step 100), and a process of updating the counter for the same time data is performed (step 102). After this, a majority decision is made based on the counter value. That is, it is determined whether the same counter value exceeds the majority (step 104). If the majority vote is not established, the processing in this routine is terminated, and if the majority vote is established, the system time is updated based on the counter value established by the majority vote (step 106).

Further, the time check correction process is performed according to the time check correction program, and the specific process is performed according to the flowchart shown in FIG. That is, to determine whether the system time has been updated normally, it is checked whether a majority of the computers in the entire system are operating (step 200). When the system time has been updated normally, the difference between the system time and the time generated by the time generation unit 26 is determined, and it is determined whether the deviation is within a set value, for example, ±1 count (step 202 ). When the deviation exceeds the set value three times in a row, the generated time is corrected to the system time (steps 204, 206), and the corrected time data is sent out (step 208). On the other hand, when the deviation is within the set value, the generated time data is sent out as is. Further, when performing time correction, the correction range is ±1 count, and the correction is performed in the direction of reducing the time difference.

In the above configuration, one time update cycle is set to 3 seconds, time data is exchanged between computers 10 and 12.14, and the system time is set to 9:59:57 (counter value 11999
The updating process from (corresponding to the counter value 12000) to 10:00 minutes 0 seconds (corresponding to the counter value 12000) will be explained with reference to FIG.

First, when the computer 10 sends the generated time data of 10:00 minutes and 0 seconds to the transmission path 16, this time data is transmitted to the computers 12 and 14 and then received by the computer 10. At this time, the computer 10 checks whether the difference with the system time counter is within ±1 count, and if there is no abnormality in the time, the counter value of the time counter is 1.
Output 2000 data. At this time, the count value of the computer 10 is 12000, but the count value of the other computers 12゜14
Since the count value of is 11999, each computer 10,
The system time on 12.14 is maintained at 9:59:57.

Next, when the time data of 0 hours, 0 minutes, and 0 seconds is output from the computer 12 to the transmission path 6, this time data is sequentially transmitted to the computer 1.
Entered at 0.14. At this time, the time data of computers 10.12 match, and this time data becomes the majority, establishing a majority vote, so the system time of each computer 10, 12.14 changes from 9:59:57 to 10:00. Updated every second.

After this, when time data of 10:00 minutes and O seconds is sent from the computer 14, this data is received by each computer.

Next, when applying the present invention to a train operation management system, a station communication station is installed at each station, each station communication station is connected via a broadcast transmission path, and information is exchanged between each station communication station. will be carried out. The departure route control in this case is performed according to the flowchart shown in FIG.

First, when implementing this control, time data is exchanged between each station communication station, and signals from train position detection sensors placed at each station are input to each station communication station. Then, at each station, based on the input time data, the time data in which a majority of the matching times are included is selected as the system time, and the time matching the system time (-softening time) is set (step 300). Thereafter, using the stopping train number as key information, the departure time of the train is extracted from the timetable information, and the departure time is compared with the matched time (step 302). Thereafter, based on the comparison result, it is determined whether or not route control of the train is required (step 304).
). In this case, if the train has stopped for a certain period of time or more and the softening time has passed the departure time, it is determined that route control is necessary and route control is executed. If route control is required, locking conditions are checked and it is determined whether there is a competing train on the route of the train in question (steps 306, 308). If there are competing trains, the priority train is determined based on the timetable information, and if another train has priority, processing is executed to give priority to the other train. executes processing for the vehicle. In this case, a departure route output request is output for the vehicle in question, and the processing in this routine ends (step 314).In addition, in this processing, all trains stopping at the station are processed in the same manner according to the priority train order. The process is executed repeatedly.

Next, when generating train arrival time records, processing at each station communication station is performed according to the flowchart shown in FIG.

First, the position of the train is checked based on the output of the train position sensor (step 400), and it is determined whether the train is accommodated on the platform (step 400). It is determined whether the train is accommodated on the platform (step 402). If the train is accommodated on the platform, a track arrival time record is created using the platform accommodation conditions (steps 404, 406, 408). In this case, it is determined whether or not an arrival record has been created for the train accommodated on the platform, and if an arrival record has not been created, the generated time data is retrieved, and the train's arrival time record is created according to this time data. . At this time, the time data used as the arrival time is the time data generated according to the matching time. As described above, in this embodiment, when performing train departure route control and creating train arrival time records, time data whose deviation from the matching time is within the set value is used as time data.
It becomes possible to operate trains according to the train schedule, and it is also possible to create accurate train arrival time records.

Further, the scope of the present invention is applicable to a distributed system composed of a plurality of computers in which the processing time of each computer must be kept within a certain tolerance range as a whole system. It is also possible to apply the present invention to management of terminal input actual times in an OA system consisting of a host computer having a database and a plurality of terminal workstations, and to a control timing matching method in a distributed FA system.

〔Effect of the invention〕

As described above, according to the present invention, since the time is corrected according to the system time based on each element constituting the system, autonomy regarding time for each element can be ensured. In addition, since the time data at each station can be corrected to match the system time, it is possible to operate trains in accordance with the train schedule, and to create accurate records regarding train operations. can.

[Brief explanation of drawings]

Fig. 2 is a flowchart for explaining the majority decision processing, Fig. 3 is a flowchart for explaining the time check correction processing, and Fig. 4 is a block diagram showing an embodiment of the present invention. FIG. 5 is a flowchart for explaining the departure route control in the train operation management system, and FIG. 6 is a flowchart for explaining the arrival record processing in the train operation management system. 10.12.14...Computer, 16...Broadcast transmission line. 18... Time data receiving section, 20... Majority decision section. 22... System time selection section, 24... Time update section. 26... Time generation unit, 28... Time check correction unit. 30...Time data transmitter.

Claims (1)

[Claims] 1. In a distributed system in which a plurality of distributed communication stations are connected to each other via an information transmission path, each communication station generates its own time and also transmits information regarding the generated time. The system time is transmitted sequentially to other communication stations, and the system time that becomes the system reference time is selected by majority vote from among the time information of all communication stations, and each communication station compares the system time with its own time. A time synchronization method for distributed systems in which when the deviation exceeds a set value, the local station's time is corrected to the system time, and a new time is generated based on the corrected time. 2. In a distributed system in which a plurality of distributed computers are connected to each other via an information transmission path, each computer generates a time and sequentially transmits information regarding the generated time to other computers. The system time that will be the reference time is selected by majority vote from among the time information of all computers, and each computer compares the system time and the generated time, and if the deviation exceeds the set value, the generated time is changed. A time matching method for distributed systems that corrects the system time and generates a new time based on the corrected time. 3. Connect multiple distributed station communication stations to each other via information communication channels, and each station communication station generates its own time and sequentially transmits information about the generated time to other station communication stations. The system time that is transmitted and becomes the system reference time is selected by majority vote from among the time information of all station communication stations, and each station communication station compares the system time with its own time and calculates the deviation. When the set value is exceeded, the local station's time is corrected to the system time, a new time that matches the system time is generated based on the corrected time, and trains are operated at each station according to the local station's time and the timetable schedule. Train operation management system.