JP2011053978A - Machine-to-machine communication system using asynchronous multicast middleware - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a machine-to-machine communication system using asynchronous multicast middleware which autonomously achieves an inter-group communication function only by many client apparatuses without using a server. <P>SOLUTION: At least one or more machines 100 such as industrial robots are connected to a network 200. In addition to a data communication I/F 10 and an application program 30, each machine 100 includes a middleware program 20 which can achieve machine-to-machine type asynchronous multicast. The middleware program 20 includes: a logical channel management function 21 between middleware programs 20; a logical channel data distribution function 22 for distributing communication data to a segment 210 in which the machine 100 belonging to the logical channel exists; and a data screening function 23 to achieve machine-to-machine type asynchronous multicast communication. <P>COPYRIGHT: (C)2011,JPO&INPIT

Description

  The present invention uses a network system constructed between machine equipment such as industrial robots on the manufacturing site, and exchanges control data and measurement data in cooperation with each other without preparing a server. The present invention relates to a communication system between machine facilities and a communication method.

  In the past, machinery and equipment such as industrial robots that were introduced at manufacturing sites were operating in a stand-alone form, but recently, in order to further increase efficiency at manufacturing sites and the like, A network system has been built, and using the network communication function, control data such as work instruction data and work report data, and measurement data such as work environment data are exchanged for efficient management of work processes and between machine equipment. Productivity is being improved through collaboration.

  When constructing a network communication system between mechanical facilities, the problem is that the components of the network are likely to fluctuate dynamically. If the network components are fixed and do not fluctuate, it is sufficient to use a so-called unicast communication method in which each device retains address information allocated once and transmits / receives data by individually specifying a communication partner. However, control of machinery and equipment such as industrial robots requires rearrangement of machinery and equipment such as industrial robots in a short period of time, and rearrangement of control programs that provide manufacturing work procedures and changes in manufacturing instruction data. The network component is easy to change. Network construction under such an environment is not easy to stably operate for a long period of time because the communication partner cannot be fixed.

  In the prior art, what is assumed as a technique for constructing a network under the above-described situation is that there are a large number of client devices without using an external server and utilizing an IP multicast function provided in a router which is a network resource. There is a method to realize the inter-group communication function autonomously with only mechanical equipment. IP multicast is a method of casting that is data exchange on a network, and is a method of transmitting data having the same contents to a plurality of specific parties who have announced participation. IP multicast is a method for transmitting data to a special IP address called a multicast address, assuming a TCP / IP network. Routers play an important role in IP multicast. Data is automatically replicated and distributed according to the number of distribution destinations by a router in the middle of the network. Since data is efficiently replicated by a router or the like at a location where the communication bus branches, even large-capacity data such as voice and video can be distributed efficiently and without compressing the communication bandwidth of the line.

However, the first method has the following problems.
First, there is a problem that the operating rate of the mechanical equipment is likely to decrease. In the first method, all management information including the address information of each machine is managed collectively by the server. Therefore, when the server stops due to a failure, or when maintenance of the server is performed Since the server does not function effectively, the entire system stops functioning eventually, and machine equipment such as industrial robots cannot be operated. In factories and the like, there is no problem with mechanical equipment such as industrial robots, but it is not permitted that the entire system is stopped due to server failure or maintenance.

  The second problem is that a server is required when networking machine equipment by the first method. In an environment where the robots participating in the network are frequently changed according to demand, change management on the server, system construction, and the like become complicated, and thus a network system cannot be constructed easily.

  The third problem is that there remains a possibility that the scalability requirement cannot be satisfied. In the first method, since all packet transmission and reception are concentrated on the server, the server processing becomes heavy when the load is concentrated, and the load cannot be distributed well to resources in the network system, improving scalability. do not do.

Next, the second method also has the following problem.
First, in order to realize IP multicast, there is a problem that all devices in the network including communication devices such as routers must be compatible with IP multicast. In other words, even when an existing router is already installed and a network is constructed, it is necessary to modify the router 220 to be compatible with IP multicast.

  The second problem is communication reliability. Problems such as packet loss may occur in network communication. However, in the IP multicast method, a retransmission function for dealing with packet loss or the like is not included in the basic function, and retransmission processing at the time of packet loss is individually implemented, and the implementation is not easy.

  Thirdly, there is a problem that it is impossible to implement advanced functions according to site requirements, such as packet delay processing. In mechanical equipment such as industrial robots at production sites such as factories, temporary detachment can frequently occur due to recombination. When returning to the network system after temporarily leaving, it is necessary to delay the packet transmitted during that time to the device, but the packet cannot be delayed by IP multicast. For this reason, when a machine facility is temporarily disconnected, the entire system is stopped and suspended until the machine facility returns to the network system, or a packet related to the case when the machine facility is temporarily disconnected or returned is retransmitted. It is necessary to implement packet resend processing to do.

  In view of the above problems, the present invention basically provides a communication system between machine facilities using asynchronous multicast middleware that autonomously realizes an inter-group communication function using only a large number of client devices without using a server. The purpose is to do.

In order to achieve the above object, a communication system between machine facilities using the asynchronous multicast middleware of the present invention, in two or more machine facilities, a network resource serving as data communication means between the machine facilities, an application program, A machine-to-machine communication system that is equipped with a middleware program that enables machine-to-machine asynchronous multicast and that transmits and receives data between the machine equipments belonging to a logical channel constructed between the machine equipments,
All of the middleware programs include a logical channel owner management table that stores data relating to logical channel owners that exist for each logical channel. In the middleware program that is the logical channel owner, A logical channel that is a node to which the node belongs, a logical channel that stores data relating to the Subscriber, and a Subscriber management table, and the middleware program that has transmission data and becomes a logical channel and Publisher searches the logical channel and the owner management table to search for the logical channel Performing logical channel data transmission / reception processing for transmitting the transmission data to the channel owner node, and the middleware program as the logical channel owner Machine-to-machine asynchronous multicast communication is enabled by executing logical channel data collection and distribution processing for distributing the transmission data to a registration destination node using a physical channel / Subscriber management table. .

  In the communication system between mechanical facilities having the above-described configuration, in addition to data transmission / reception using a logical channel, various processes as described below are possible.

  The first is processing for constructing a logical channel in a communication system between machine facilities. If the logical channel owner related to the search does not exist as a result of the search of the logical channel owner management table, the middleware program of the node that wants to become the logical channel owner generates a logical channel object and identifies itself as the logical channel owner. Register in the logical channel owner management table managed by itself as a channel owner node, and broadcast to the middleware program of other nodes broadcast data declaring that it is the logical channel owner. It is assumed that the channel is built.

  The second is a process of joining an existing logical channel in the communication system between machine facilities. The middleware program of the node that wants to participate searches its own logical channel owner management table, unicasts data indicating a participation request to the searched logical channel owner, and from the logical channel owner. When data indicating participation permission is received, the node of the logical channel owner is registered in the logical channel owner management table managed by itself, and the logical channel owner manages itself when issuing the participation permission. The process of registering the node that has transmitted the data indicating the participation request in the logical channel / Subscriber management table is performed.

  The third is a process in which a node leaves a logical channel already participating in the communication system between machine facilities. The middleware program of the node desiring to leave searches its own logical channel owner management table, unicasts data indicating a withdrawal request to the searched logical channel owner, and from the logical channel owner. When data indicating leave permission is received, the logical channel owner node is deleted from the logical channel owner management table managed by itself, and the logical channel owner manages it when issuing the leave permission. The processing is to delete the node that has transmitted the data indicating the withdrawal request from the logical channel / Subscriber management table.

  The fourth is processing in which a node temporarily leaves a logical channel that has already participated in a communication system between machine facilities. The middleware program of the node that desires temporary processing searches its logical channel owner management table, unicasts data indicating a temporary withdrawal request to the searched logical channel owner, and When receiving data indicating permission to temporarily leave from the owner, the logical channel owner node is treated as being temporarily stopped in the logical channel owner management table managed by the owner, and the logical channel owner issues the leave permission. At this time, the node that has transmitted the data indicating the temporary withdrawal request from the logical channel / subscriber management table managed by itself is treated as a suspended state.

Fifth, in the communication system between machine facilities, is processing when the machine device leaves the network itself. When the middleware program of the node of the machine facility to be detached searches the logical channel owner table and detects whether there is a logical channel object whose own node is the logical channel owner, The channel owner transmitted from the node corresponding to the logical channel / owner substitution request is transmitted to all Subscriber nodes registered in the logical channel / subscriber management table. Receiving the compatibility information, selecting one of the subscriber nodes judged to be optimal, executing the logical channel / owner transfer process for transmitting all the logical channel / subscriber management table data, and executing the logical channel owner Searches the management table and himself / herself This is a process of detecting a logical channel participating as an iber node and executing the leaving or temporary leaving process from the logical channel.
In addition, the process step which implement | achieves the communication system between machine facilities of this invention can be provided with the form of a program.

  According to the communication system between machine facilities using the asynchronous multicast middleware according to the present invention, a network resource is mounted on a machine facility such as an industrial robot, and an application has a fixed address when performing communication via the network. Without specifying information, the logical “logical channel” managed by the middleware is dynamically specified, leaving communication to the middleware, and machine-to-machine communication is executed asynchronously between the middleware. If distribution is performed for all the segments belonging to the logical channel, it is delivered to the application via the middleware of the node belonging to the logical channel, and desired communication can be performed. Further, since a specific server is not required and the machine-to-machine method is used, it is possible to avoid the entire system from being stopped due to a server failure.

  Hereinafter, a communication system between machine facilities and a communication method between machine facilities of the present invention will be described in detail based on embodiments shown in the accompanying drawings. The present invention is not limited to these examples.

  A communication system 100 between machine facilities using the asynchronous multicast middleware according to the first embodiment of the present invention will be described. In the following configuration example, it is assumed that an industrial robot is assumed as the mechanical equipment 100.

  FIG. 1 is a diagram schematically illustrating a configuration example of a communication system between machine facilities according to the first embodiment of the present invention. As shown in FIG. 1, in the communication system between machine facilities of the present invention, a network system is constructed by connecting at least one or more machine facilities 100 via a network resource 200 serving as data communication means. Although various network configurations are possible, in the example of FIG. 1, only one segment of the network 200 is shown very simply, and only three of the machine equipment 100a, the machine equipment 100b, and the machine equipment 100c are shown. Yes. In the drawing, suffixes a, b, c, etc. are appropriately added to the numbers of the respective components.

  In the communication system between machine facilities of the present invention, communication data is reliably distributed to one or a plurality of destinations to be distributed by middleware. In other words, in the present invention, the application always issues a request using the unicast method when making a data distribution request regardless of whether the distribution destination is a plurality of destinations or a single destination. The request is internally converted into a broadcast method or a unicast method for each destination, and distribution processing is performed. The middleware 20 dynamically manages the destination for each logical channel, and there is no need to grasp the address information of the distribution partner at the application level. Therefore, in the environment where the destination of data collection and delivery changes dynamically due to changes according to the production plan or temporary withdrawal due to malfunction, the machine equipment 100 manages all the configuration changes autonomously and is transparent to the application. Therefore, the communication method is suitable in such an environment.

  Here, as mentioned in the column of the prior art and the problem to be solved, there is a problem that the retransmission processing at the time of packet loss cannot be performed in the conventional IP multicast method, such as when the communication system between machine facilities is moving In a use environment where the user leaves the network for a short time and is connected to the network, a situation in which data cannot be received frequently occurs. On the other hand, in the communication system between machine facilities of the present invention, it is possible to realize the multicast system without affecting the network resources only by introducing the middleware 20, and appropriately retransmit or delay the transmission when there is a packet loss problem as will be described later. Processing can be performed and the problem of data reception loss can be solved.

  In the communication system 100 between machine facilities of this invention, data communication is performed via the logic channel built between each machine facility. A software object called a “logical channel object” is generated and stored in each machine facility, and data communication is performed between the same logical channel objects to enable selective data communication between the machine facilities.

  The logical channel object is a software object generated in units of logical channels such as channel m and channel n, and implements various tables (logical channel / owner management table, logical channel / subscriber management table) and delay distribution described later. Therefore, a temporary queue mechanism is included.

  In this way, there can be multiple logical channels in a network, and these logical channel objects can exist on separate machines, and multiple logical channel objects can be on a single machine. It can also exist. The data stored in the temporary queue can be present on the memory, on the non-volatile medium, or both. If it exists on a non-volatile medium, even if the mechanical device is reset, data stored in the temporary queue is not lost, and data distribution can be continued after recovery from reset.

  For each logical channel, one logical channel object in the network is positioned as the logical channel owner. A logical channel object has only one owner in the network for each channel. The logical channel owner is dynamically determined by the logical channel creation process (Build), and the role of the logical channel owner and its logical channel object content can be delegated dynamically to other machine-to-machine communication systems. Is possible. Data from Publisher is received by the owner, and the logical channel owner distributes it to the registered subscriber of the logical channel. One node may be a plurality of logical channel owners, and conversely, one node may not perform the logical channel owner function at all. In order to avoid a situation where a plurality of logical channel owners exist for the same logical channel in the network, a protocol as described later is adopted.

The configuration of each mechanical facility 100 will be described with reference to FIG.
As shown in FIG. 1, the machine facility 100 includes a data communication interface 10, a middleware program 20, an application program 30, and a machine facility function 40. The operating system program is not shown. The operating system may be an operating system installed in the machine facility 100, but is preferably a network compatible one.

  The data communication interface 10 also includes physical modules necessary for network connection such as connectors and network cards. For example, the data communication interface 10 may be a conventional technology such as Ethernet widely used in a LAN or the like, and an interface technology developed in the future can also be applied.

  The application program 30 is a program for operating the original mechanical facility function 40 of the mechanical facility 100. In the present invention, the configuration of the application program 30 is not particularly limited, and the language is not limited. However, the interface for data exchange with the middleware program 20 needs to match. For example, it can be dealt with by leaving the application program 30 as it is and making the interface with the application program 30 on the newly introduced middleware program 20 side.

  The machine facility function 40 is an original function of each machine facility, and is a function provided by the robot if it is an industrial robot. There may be various functions depending on the purpose and configuration of the mechanical equipment 100. The machine facility function 40 is driven by the application program 30.

  The middleware program 20 is a middleware program that enables machine-to-machine asynchronous multicast. When the middleware program 20 receives the communication data from the application 30, the middleware program 20 transmits data communication to the middleware 20 of the other machine equipment 100 via the network with the logical channel identifier designated by the application. . In addition, when communication data from the middleware 20 of another machine equipment 100 is received via the data communication interface 10, the role of passing the received communication data to the application 30 installed in the machine equipment 100 is provided. Fulfill.

The middleware program 20 includes a logical channel management function 21, a logical channel data transmission / reception function 22, and a logical channel data collection / delivery function 23.
First, the logical channel management function 21 will be described.
The logical channel management function 21 is a function provided by the middleware program. The logical channel management function 21 maintains the logical channel / owner management table 211 and the logical channel / subscriber management table 212, and receives the data transmission request and its data passed from the application layer. Send to logical channel owner of logical channel. If the own node is the logical channel owner, the data received through the network is distributed to the subscriber (s) registered in the logical channel. In this process, a delivery confirmation and a process of temporarily delaying the data when the machine facility is temporarily removed from the network and the machine facility returns to the network are also performed. For example, by exchanging information with machine equipment in the network, a table having the following information is constructed and maintained and managed dynamically at all times.

As shown in [Table 1], the logical channel owner management table 211 is a table composed of a plurality of {logical channel name, destination address, status} entries.

  The logical channel / subscriber management table 212 is a table for managing information on logical channels / subscribers belonging to each logical channel, as shown in [Table 2]. When the logical channel owner receives data from the logical channel / Publisher, the logical channel owner searches the table and distributes data to all the Subscribers belonging to the corresponding logical channel. The owner of each logical channel manages the Subscriber list of the logical channel that it owns in this table. Although it is possible to manage a subscriber list of logical channels that are not managed by itself, it is not essential.

  Next, when the logical channel data transmission / reception function 22 receives a transmission request from the application layer, the logical channel / data transmission function searches the logical channel / owner management table managed by the logical channel management function and acquires the acquired logical channel. • Send data to the logical channel owner using the owner's destination address. At this time, by adding the transmission sequence number in addition to the transmission data and performing transmission, the data is reliably delivered to the logical channel owner. When the logical channel / data transmission / reception function of the logical channel / Publisher receives an Ack (Acknowledge) message, the logical channel owner regards the data as having been received by the logical channel owner and discards the data. If an Ack (Acknowledge) message is not received within a certain time, the data is retransmitted on the assumption that the data has not been received by the logical channel owner.

  Next, the logical channel data collection and distribution function 23 is a function used when the logical channel owner receives or distributes data. The logical channel owner receives the data transmitted from the logical channel / Publisher using this function. At that time, the transmission sequence number included in the transmission data is read, and Ack (Acknowledge) data including the sequence number is returned to the logical channel / Publisher.

Based on the above various functions, various data communication processes are executed in the communication data network constructed between the machine facilities 100.
Hereinafter, basic processing in the communication system 100 between machine facilities of the present invention will be described.

  First, the logical channel creation process (Build) will be described, and then the logical channel participation procedure (Join) in which another node participates in the created logical channel, and the node leaves the participating logical channel. Logical channel leave processing (Leave), logical channel temporary leave processing (Suspend) in which a node temporarily leaves a participating logical channel will be described, and then, transmission processing to the logical channel as data communication using the logical channel (Publish) and reception processing (Subscribe) from a logical channel will be described.

<Logical channel creation processing (Build)>
The data flow of the logical channel creation processing will be described with reference to the data flow diagrams of FIGS. 2 to 4 and the flowchart of FIG.

  First, as shown in the data flow of (1) in FIG. 2, in this example, when the machine facility 100a receives a channel creation request from the application program 30 (step S1 in FIG. 16), the logical channel management function 21 performs the logical channel management function 21. The channel owner management table 211 is searched to check whether the logical channel already exists (step S2 in FIG. 16).

  Here, if the logical channel already exists in the logical channel owner management table 211 (step S2 in FIG. 16: Y), an error is returned to the application program 30 (to the end of FIG. 16). If no entry exists in the logical channel owner management table 211 (step S2: N in FIG. 16), [Build Channel Request] is broadcasted to the same LAN segment as shown in (1) of FIG. (FIG. 16, step S3). In this example, [Build Channel Request] is received for the machine equipment 100b and the machine equipment 100c.

  There are the following two cases in response to a logical channel creation request transmitted from the machine facility 100a to each node of the network.

The first case is when there is a logical channel owner on the network.
If the logical channel management function 21 of each machine facility in the same LAN segment that has received the [Build Channel Request] message is the logical channel owner of the logical channel, as shown in the flow (2) of FIG. Then, a [Channel Exist] message is returned by broadcast transmission. In this configuration example, the machine facility 100b is the logical channel owner of the logical channel, and returns a [Channel Exist] message by sending a broadcast message.

  On the other hand, other nodes that have received the [Channel Exist] message by broadcast register the logical channel in the logical channel owner management table 211 managed by their own logical channel management function 21 so that the logical channel is treated as existing. In this configuration example, the mechanical equipment 100c receives the [Build Channel Request] message from the mechanical equipment 100a, and then adds the channel to the logical channel / owner management table 211 according to the [Channel Exist] message received from the mechanical equipment 100b. Register as existing. The entries in this table are deleted after a certain period (valid only for a certain period).

  When the logical channel management function 21a of the machine facility 100a receives the [Channel Exist] message (step S4: Y in FIG. 16), the logical channel management function 21a changes the [Channel Exist] message as shown in the data flow in (3) of FIG. In response, an ACK message is returned (step S6 in FIG. 16). In this configuration example, the machine equipment 100a that has transmitted the [Build Channel Request] message returns an ACK message in response to the [Channel Exist] message returned from the machine equipment 100b, and the machine equipment 100b receives the ACK message. .

The second case is when there is no logical channel owner on the network.
When the [Channel Exist] message is not received from another node on the network within a certain time (step S4: N in FIG. 3), the machine equipment 100a determines that the currently active logical channel does not exist in the segment. Then, a logical channel object is generated and registered in the logical channel owner management table 211 that manages itself as a logical channel owner (step S5 in FIG. 3), as shown in the data flow of (4) in FIG. In addition, a [Channel Already Exist] message is transmitted to the other nodes in the segment to ensure that it is the logical channel owner (step S5 in FIG. 3).

As shown in the data flow of (5) in FIG. 4, the logical channel management function 21 of the machine facility 100a sends information on the channel for which it is the logical channel owner to a segment as a [Channel Exist] message at regular intervals. Broadcast message transmission.
The logical channel creation process (Build) is executed by the above procedure.
Thereafter, as will be described later, data communication using this logical channel becomes possible.

<Logical channel participation processing (Join)>
Next, a logical channel participation process (Join) in which another node participates in a logical channel that has already been created will be described.
The data flow of the logical channel participation process will be described with reference to the data flow diagrams of FIGS. 5 to 6 and the flowchart of FIG. As an example, a process in which the machine facility 100a joins a logical channel in which the machine facility 100b is a logical channel owner will be described.

  When the machine facility 100a receives a logical channel participation request from the application program 30 (step S1 in FIG. 17), the logical channel management function 21 first searches the logical channel owner management table 211. If an entry already exists in the logical channel owner management table 211 (step S2 in FIG. 17: Y), a [Join Channel Request] is transmitted as a unicast message according to the logical channel owner destination information (FIG. 17). 17 step S3). In FIG. 5, illustration in the case of unicast communication when an entry exists in the logical channel / owner management table 211 is omitted.

  If no entry exists in the logical channel / owner management table 211 (step S2 in FIG. 17: N), as shown in (1) of FIG. 5, [Join Channel Request] is broadcasted to the same LAN segment. (FIG. 17 step S4).

  Instead of sending a broadcast message for [Join Channel Request], send [Search Channel Request] by broadcast, and wait for the [Notify Channel Owner] message to be returned from this channel owner. [Join Channel Request] may be sent to the channel owner as a unicast message.

  If the logical channel management function 21 in the machine device 100 in the segment that has received the [Join Channel Request] message is the logical channel owner, it returns a [Join Accept] message by sending a Broadcast message. In this example, as shown in (2) of FIG. 6, the machine equipment 100b sends back a [Join Accept] message by sending a Broadcast message, and in the logical channel / Subscriber management table 212 managed by itself in parallel, [Join Channel Request The address of the machine 100a that has transmitted the message is registered in the Subscriber / address field.

  If the machine facility 100a does not receive a [Join Accept] message from another node within a certain time (step S5: N in FIG. 4), it is determined that the logical channel owner does not exist in the same segment, and an error is detected. (Step S8 in FIG. 17).

  When the machine facility 100a receives a [Join Accept] message from another node within a predetermined time as shown in (2) of FIG. 6 (step S5: Y in FIG. 17), the logical channel management function 21 The destination address of the logical channel owner is acquired from the received message and registered in the logical channel owner management table 211 managed by itself, and the request processing to the application is returned to the application program 30 (step S6 in FIG. 17). .

  In a node other than the machine equipment 100a (here, the machine equipment 100b) that has received the [Join Accept] message, the logical channel management function 21 registers the destination address of the logical channel owner in the logical channel owner management table 211. Entries in this table expire after a certain period and are deleted (valid only for a certain period).

The machine equipment 100a that has received the [Join Accept] message returns an ACK message to the channel owner that has transmitted the [Join Accept] message as shown in FIG. 6 (4) (step S7 in FIG. 17).
The logical channel participation process (Join) is executed by the above procedure.
Thereafter, as will be described later, data communication using this logical channel becomes possible.

<Logical channel leave processing (Leave)>
Next, a logical channel leaving process (Leave) for leaving a logical channel in which a certain node has already joined will be described.
The data flow of the logical channel leaving process will be described with reference to the data flow diagrams of FIGS. 7 to 8 and the flowchart of FIG. As an example, a process will be described in which the machine facility 100a leaves the logical channel for which the machine facility 100b is the logical channel owner.

  In the machine facility 100a, the logical channel management function 21 that has received the logical channel withdrawal request from the application program 30 (step S1 in FIG. 18) first searches the logical channel owner table 211 managed by itself. When the logical channel information exists in the logical channel owner table 211 (step S2 in FIG. 18: Y), [Leave Channel Request] is directly unicasted to the logical channel owner indicated by the entry information (step 18 in FIG. 18). S6).

  In FIG. 7, the illustration in the case of unicast communication in the case where an entry exists in the logical channel / owner management table 211 is omitted.

  On the other hand, if the channel information does not exist in the logical channel owner table 211 (step S2: N in FIG. 18), [Search Channel Request] is broadcast to the same LAN segment as shown in (1) of FIG. (Step S3 in FIG. 18) and waits for a reply of the [Notify Channel Owner] message from the logical channel owner.

As shown in FIG. 8 (2), when the machine facility 100a that broadcasts [Search Channel Request] receives a [Notify Channel Owner] message from the logical channel owner (step S4: Y in FIG. 18), FIG. As shown in (3), channel owner information is reflected in the logical channel owner table 211, and [Leave Channel Request] is unicasted (step S6 in FIG. 18).
If the [Notify Channel Owner] message does not return from the logical channel owner within a certain period of time (step S4: N in FIG. 6), [Search Channel Request] is retransmitted and several attempts are made, and errors are counted. If repeated, an error is returned to the application (step S5 in FIG. 18).

Upon receiving the [Leave Channel Request] message, the logical channel management function 21 of the machine facility 100b that is the logical channel owner in the same LAN segment deletes the entry of the leave processing requester from the logical channel / Subscriber table 212. As shown in FIG. 8 (4), an [Accept leave request] message is returned to the machine equipment 100a as a unicast message.
If the machine equipment 100a receives the [Accept leave request] message (step S7: Y in FIG. 18), it confirms that the logical channel has been successfully removed (step S8 in FIG. 18). If no [Accept leave request] message is received (step S7: N in FIG. 18), an error is returned to the application 40 (step S9 in FIG. 18).

With the above procedure, the process of leaving the logical channel (Leave) is executed.
Thereafter, as will be described later, the machine facility 100b is not involved in data communication using this logical channel.

<Logical channel temporary disconnection processing (Suspend)>
Next, logical channel temporary leave processing (Suspend) in which a certain node temporarily leaves a logical channel in which a node has already joined will be described.
The data flow of the logical channel temporary disconnection processing will be described with reference to the data flow diagrams of FIGS. 9 to 10 and the flowchart of FIG. As an example, a process will be described in which the machine facility 100a temporarily leaves the logical channel for which the machine facility 100b is the logical channel owner.

  The basic process of the logical channel temporary leaving process (Suspend) is the same as the logical channel leaving process (Leave). The difference is that when a leave request is received on the channel owner side, the sender's entry is deleted from the logical channel / subscriber table 212, but when a suspend request is received, the sender entry in the logical channel / subscriber table 212 is deleted. The status is changed to “Receiving stopped”.

  In the machine facility 100a, the logical channel management function 21 that has received the logical channel temporary withdrawal request from the application program 30 (step S1 in FIG. 19) first searches the logical channel owner table 211 managed by itself. If the logical channel information exists in the logical channel owner table 211 (step S2 in FIG. 19: Y), [Suspend Channel Request] is directly unicasted to the logical channel owner indicated by the entry information (FIG. 19). Step S6). In FIG. 9, the illustration in the case of unicast communication in the case where an entry exists in the logical channel / owner management table 211 is omitted.

  On the other hand, when the channel information does not exist in the logical channel owner table 211 (step S2: N in FIG. 19), as shown in (1) of FIG. 9, [Search Channel Request] is broadcasted to the same LAN segment. (Step S3 in FIG. 19) and waits for a reply of the [Notify Channel Owner] message from the logical channel owner (Step S4 in FIG. 19).

When the machine facility 100a that broadcasts [Search Channel Request] receives the [Notify Channel Owner] message from the logical channel owner as shown in FIG. 10 (2) (step S4: Y in FIG. 19), FIG. As shown in (3), the channel owner information is reflected in the logical channel owner table 211, and [Suspend Channel Request] is unicasted (step S6 in FIG. 19).
If the [Notify Channel Owner] message does not return from the logical channel owner within a certain period of time (step S4: N in FIG. 19), [Search Channel Request] is retransmitted and several attempts are made, and errors are counted. If repeated, an error is returned to the application (step S5 in FIG. 19).

The logical channel management function 21 of the machine facility 100b that is the logical channel owner in the same LAN segment that has received the [Suspend Channel Request] message receives the status in the entry of the Suspend processing requester from the logical channel / Subscriber table 212. The column is changed to “reception stopped”, and as shown in (4) of FIG. 8, the [Accept suspend request] message is returned to the machine equipment 100a as the message sender by unicast transmission.
When the machine facility 100a receives the [Accept suspend request] message (step S7: Y in FIG. 19), it confirms that the logical channel has been successfully disconnected (step S8 in FIG. 18). If the [Accept suspend request] message is not received (step S7: N in FIG. 18), an error is returned to the application 40 (step S9 in FIG. 18).

Through the above procedure, the temporary leave processing (Suspend) from the logical channel is executed.
Thereafter, as will be described later, the machine facility 100b is temporarily not involved in data communication using this logical channel.

<Data transmission / reception processing using logical channels (Publish / Subscribe)>
Next, a data flow of data transmission / reception using the constructed logical channel will be described.
The process (Publish / Subscribe) for transmitting data between the mechanical facilities 100 using the constructed logical channel is described with reference to the flow charts of FIGS. 11 to 12 and the flowcharts of FIGS. I will explain. As an example, a process when the machine equipment 100a performs data transmission using a logical channel in which the machine equipment 100b is a logical channel owner will be described.

  In the machine facility 100a, the logical channel management function 21 that has received a logical channel / data transmission request from the application program 30 (step S1 in FIG. 20) searches the logical channel / owner management table 211. If there is no channel owner information regarding this channel in the logical channel owner management table 211 (step S2 in FIG. 20: N), the logical channel management function 21 returns an error to the application program 30 (step S3 in FIG. 20). At this time, the application program 30 can re-execute the above [Join Channel Request] process.

  If there is an entry in the table (step S2 in FIG. 10: Y), as shown in FIG. 11 (1), a [Publish Channel data] message is unicasted according to the logical channel owner destination information (FIG. 11). 10 step S4). The application program 30 is informed that the processing is successful at this point. In this example, it is assumed that the machine facility 100b is registered in the logical channel owner management table 211. The machine equipment 100a is called a logical channel / Publisher related to the data. When transmitting a [Publish Channel data] message, the machine equipment 100a that performs data communication sends a sender sequence number that is set and managed in units of logical channels. The value of the sender sequence number is incremented by 1 for each transmitted message.

  Next, as shown in (2) of FIG. 11, the mechanical equipment 100b that is the channel owner that has received the [Publish Channel data] message uses the logical channel data collection and distribution function 23 to place a temporary queue for the logical channel object. The received transmission message is stored, the sender sequence number included in the received data is read, and an Ack (Acknowledge) message including the sequence number information is returned to the machine facility 100a which is the logical channel / Publisher.

  If the machine equipment 100a which is the logical channel / Publisher cannot receive the Ack message from the machine equipment 100b which is the logical channel owner within a certain time (step S5: N in FIG. 20), the retransmission processing of the data is repeated several times. If it is not successful, an error is returned to the application 30 (step S6 in FIG. 20). When the machine equipment 100a which is the logical channel / Publisher can receive the Ack message from the machine equipment 100b within a predetermined time (step S5: Y in FIG. 20), the sender sequence number and the transmission processing requested message included in the message. If the sender sequence numbers match, the message is deemed to have been delivered to the logical channel owner, and the data is deemed to have been processed and is erased from the buffer (step S7 in FIG. 20). The logical channel management function 21 notifies the application 30 that the data transmission has been completed (step S8 in FIG. 20).

  On the other hand, the logical channel owner (in this example, the machine facility 100b) that has received the transmission message in the temporary queue (step S1: Y in FIG. 21) searches the logical channel / Subscriber management table 212, and displays (3) in FIG. As shown in the figure, the registered subscriber list addresses are obtained and managed for each channel object unit to all Subscriber addresses (in this example, only the machine facility 100c) except the transmission node (in this example, the machine facility 100a). The received data stored in the temporary queue with the assigned sequence number is transmitted as a [Deliver Channel data] message (step S2 in FIG. 21).

  Each Subscriber of the logical channel (in this example, the machine facility 100c) receives the delivered [Deliver Channel data] message (step S1 in FIG. 22), as shown in (3) of FIG. The logical channel / data transmission / reception function 23 checks the channel distribution sequence number assigned to the data, and receives the distribution sequence number of the logical channel received before receiving the data and the distribution sequence number of the channel of the latest data received. If the sequence number is a sequential number (step S2 in FIG. 22: Y), as shown in (4) of FIG. 12, [ACK Deliver Channel data] is sent to the machine equipment 100b that is the logical channel owner. ] Is sent back to the application 30c (step S4 in FIG. 22).

  If the sequence number is not a sequential number (step S2: N in FIG. 22), the last sequence number that was a sequential number and the following bitmap information (if the bit is OFF, the corresponding data has not been received, the bit is If it is ON, it indicates that the corresponding data has been received), and [NACK Deliver Channel data] is unicasted back to the machine equipment 100b that is the logical channel owner (step S4 in FIG. 22).

  The machine facility 100b, which is the logical channel owner, waits for the ACK or NACK response received from each of the subscribers that delivered the message, cannot receive ACK within a certain time (step S3: N in FIG. 21), and has not received NACK. In this case (step S4 in FIG. 21: N), data distribution is tried several times to the Subscriber, and if it fails, an error is returned to the mechanical equipment 100a (step S5 in FIG. 21).

  When the machine facility 100b that is the logical channel owner that has delivered the [Deliver Channel data] message has received an ACK from the Subscriber (step S3: Y in FIG. 21), the channel delivery sequence number included in the ACK message is set to the logical channel. Reflected in the delivered sequence number column of the relevant entry in the Subscriber management table 212 (step S6 in FIG. 21). FIG. 25A shows an example of the bit data string of the Ack message.

  When the machine facility 100b, which is the logical channel owner that delivered the [Deliver Channel data] message, receives a NACK from the Subscriber (step S4: Y in FIG. 21), the channel delivery sequence number and bitmap included in the NACK message And at the same time, the distribution data that has not been received is redistributed to this Subscriber (step S7 in FIG. 21). FIG. 25B shows an example of the bit data string of the Ack message.

  The machine equipment 100b that is the logical channel owner periodically checks the delivered sequence number fields of all applicable Subscriber entries belonging to this channel in the logical channel / Subscriber management table 212, and confirms the completion of reception by all Subscribers. The data (step S8 in FIG. 21) ends normally and is deleted from the temporary queue (step S9 in FIG. 21).

The [Deliver Channel data] message is redistributed up to a certain number of times, and if neither ACK nor NACK is received, the subscriber status is changed to suspended. Deliver undelivered messages as [Deliver Channel data] messages at regular intervals or periodically send [Send Heart Beat] messages to Subscribers whose status is suspended Then, return the status from the paused state to the active state.
Through the above procedure, data transmission / reception is performed using a logical channel.

<Processing when machine device leaves network>
Finally, the flow of processing when the machine device 100 leaves the network itself will be described with reference to the data flow diagrams of FIGS. 13 to 15 and the flowcharts of FIGS. 23 to 24.
When the machine apparatus 100 leaves the network itself, a logical channel / owner transfer process and a leave / temporary leave process from the logical channel are performed. When the logical channel management function 21 of the machine apparatus 100 receives a request for temporarily leaving or leaving the network from the application 30 (step S1 in FIG. 23), whether or not there is a channel object for which the channel owner is itself. The logical channel / owner table 211 is searched for determination (step S2 in FIG. 23).

As a result of the determination by searching the logical channel owner table 211, if there is a channel object that is the channel owner (step S2 in FIG. 23: Y), as shown in FIG. A [Channel ownership consideration] message is unicasted to all Subscribers registered in the object (step S3 in FIG. 23).
On the other hand, if there is no channel object that is the channel owner (N in FIG. 23), the process ends.

  The logical channel management function 21 of the machine facility 100b scheduled to leave the network waits for a certain period of time to return a [Channel ownership request] message from one or more Subscribers who have sent the [Channel ownership review] message (step in FIG. 23). S4). If the [Channel ownership request] message is not returned (step S4: N in FIG. 23), the [Channel ownership review] message is unicasted again to all the Subscribers, and the nodes that wish to become logical channel owners are recruited. (Return to step S3 in FIG. 23). Note that if it does not succeed after several iterations, it is treated as an error and the process ends.

  On the other hand, the Subscriber that has received the [Channel ownership review] message (step S1: Y in FIG. 24), as shown in FIG. 13 (2), displays information that becomes an index indicating that the own node is compatible with the alternative channel owner [ It is added to the “Channel ownership request” message and returned (step S2 in FIG. 24). As information serving as an index suitable for an alternative channel owner, attribute information such as the elapsed time from completion of activation and the presence / absence of a non-volatile medium for temporarily storing a delivery message and its storage capacity can be considered. In this configuration example, a [Channel ownership request] message is returned only from the machine facility 100b.

  When the [Channel ownership request] message is returned from the subscriber on the network (Step S4: Y in FIG. 23), the logical channel management function 21 of the machine facility 100b scheduled to leave the network is included in the [Channel ownership request] message. Each “channel owner suitability information” is compared, and one subscriber determined to be optimal is selected (step S5 in FIG. 23). This Subscriber is called “next channel owner”.

  The logical channel management function 21 of the machine facility 100b scheduled to leave the network transmits a [Channel ownership delegation] message to the selected “next channel owner” as shown in (3) of FIG. As shown in FIG. 23, all the distribution data and logical channel / Subscriber management table data stored in the temporary queue in the channel / object are transmitted to “next channel owner” (step S6 in FIG. 23). In this transmission, in order to pass all the information to the “next channel owner”, a delivery confirmation is performed by giving a sequence number to each transmission data, or an OS communication function having a delivery confirmation function (for example, TCP) is preferably used.

  next If the machine equipment 100c that becomes the channel owner succeeds in the delivery of the distribution data and the logical channel / Subscriber management table data (step S3: Y in FIG. 24), the Ack signal is returned as shown in (5) of FIG. (Step S4 in FIG. 24). If the data transfer is not successful even after a lapse of a certain time (step S3 in FIG. 24: N), a Nack signal is returned (step S5 in FIG. 24).

  The logical channel management function 21 of the machine facility 100b scheduled to leave the network completes the process of delivering all distribution data and logical channel / subscriber management table data (step S7: Y in FIG. 23), and then (6) in FIG. As shown in FIG. 23, the fact that the channel owner has been changed is transmitted to all Subscribers registered in the channel object (step S8 in FIG. 23). For example, it transmits the [Notify Channel Owner] message including the address of “next channel owner”. This series of steps completes the channel owner relocation work.

  If the process of passing all distribution data and logical channel / Subscriber management table data is not completed even after a certain period of time has passed, the sending node selected as the “next channel owner” is inappropriate or there is a problem in data communication. In order to select another node again, the process returns to step S5 in FIG.

  Next, the logical channel management function 21 of the machine facility 100b scheduled to leave the network acquires information on the logical channels that it participates as a Subscriber by searching the logical channel owner management table 211 and the like, and leaves each logical channel. Processing is performed (step S9 in FIG. 23). That is, [Leave Channel Request] or [Suspend Channel Request] is unicast transmitted to the channel owner of each channel.

When each logical channel leaving process in which the machine facility 100b scheduled to leave the network is completed, the machine device 100b can leave the network.
With the above procedure, the machine device 100b can leave the network itself, and even if the machine device 100b leaving the network is a logical channel owner, only the machine device 100b is not affected by the logical channel. Can be removed.

  As mentioned above, although the preferred example in the example of composition of the communication system between machine facilities of the present invention has been illustrated and explained, it is understood that various changes are possible without departing from the technical scope of the present invention. I will.

  The communication system between machine facilities of the present invention can be widely applied to a production system in which an industrial robot is introduced.

The figure which shows typically the structural example of the communication system between machine equipment which concerns on Example 1 of this invention. Diagram for explaining the data flow of logical channel creation processing (part 1) Diagram for explaining data flow of logical channel creation processing (part 2) A diagram for explaining the data flow of logical channel creation processing (part 3) Diagram for explaining the data flow of logical channel participation processing (part 1) Diagram for explaining data flow of logical channel participation processing (part 2) A diagram for explaining the data flow of logical channel leaving processing (part 1) A diagram for explaining the data flow of logical channel leaving processing (part 2) A diagram for explaining the data flow of logical channel temporary disconnection processing (part 1) Diagram for explaining data flow of logical channel temporary disconnection processing (part 2) A diagram for explaining a data flow of data transmission processing using a logical channel (No. 1) A diagram for explaining the data flow of data transmission processing using a logical channel (part 2) Diagram explaining the flow of network disconnection processing (part 1) Diagram explaining the flow of network disconnection processing (part 2) Diagram explaining the flow of network disconnection processing (part 3) Flow chart showing the flow of logical channel creation processing Flow chart showing the flow of logical channel participation processing Flowchart showing the flow of logical channel leaving processing Flow chart showing the flow of logical channel temporary disconnection processing Flow chart showing the flow of processing in the sending node that requested data transmission Flowchart showing the flow of processing at the sending node that is the logical channel owner Flow chart showing the flow of processing in the sending node that is the Subscriber of the corresponding logical channel A flowchart showing the flow of processing in a node leaving the network Flow chart showing the flow of processing in the node that will newly become the logical channel owner The figure which shows the example of the bit data sequence of an Ack message, and the example of the bit data sequence of a Nack message

DESCRIPTION OF SYMBOLS 10 Data communication interface 20 Middleware program 21 Logical channel management function 22 Logical channel data transmission / reception function 23 Logical channel data collection / delivery function 30 Application program 40 Mechanical equipment function 100 Mechanical equipment 200 Network

Claims (8)

In two or more machine facilities, a network resource that is a data communication means between the machine facilities, an application program, and a middleware program that enables machine-to-machine asynchronous multicasting, and a logic constructed between the machine facilities A communication system between machine facilities for transmitting and receiving data between the machine facilities belonging to a channel,
All of the middleware programs comprise a logical channel owner management table that stores data relating to logical channel owners that exist for each logical channel,
In the middleware program that is the logical channel owner, a logical channel that is a node belonging to the logical channel, a logical channel that stores data related to the Subscriber, a Subscriber management table,
The middleware program that has transmission data and is a logical channel Publisher executes logical channel data transmission / reception processing for searching the logical channel owner management table and transmitting the transmission data to the logical channel owner node,
The middleware program, which is the logical channel owner, executes a logical channel data collection and distribution process for distributing the transmission data to a registration destination node using the logical channel / subscriber management table. Type machine-to-machine communication system characterized by enabling asynchronous multicast communication.   If the logical channel owner related to the search does not exist as a result of the search of the logical channel owner management table, the middleware program of the node that wants to become the logical channel owner generates a logical channel object and identifies itself as the logical channel owner. Register in the logical channel owner management table managed by itself as a channel owner node, and broadcast to the middleware program of other nodes broadcast data declaring that it is the logical channel owner. The communication system according to claim 1, wherein a channel is constructed. The middleware program of a node that wants to participate in an existing logical channel searches its logical channel owner management table, unicasts data indicating a participation request to the searched logical channel owner, When data indicating permission to participate is received from the channel owner, the logical channel owner node is registered in the logical channel owner management table managed by itself,
3. The logical channel owner registers the node that has transmitted the data indicating the participation request in the logical channel / subscriber management table managed by the logical channel owner when granting the participation permission. The communication system between machine facilities described in 2. When a node leaves a logical channel that has already joined, the middleware program of the node searches its own logical channel owner management table and stores data indicating a withdrawal request for the searched logical channel owner. When unicast transmission and data indicating permission to leave are received from the logical channel owner, the logical channel owner node is deleted from the logical channel owner management table managed by itself,
4. The logical channel owner deletes the node that has transmitted the data indicating the withdrawal request from the logical channel / subscriber management table managed by itself when issuing the leave permission. The communication system between machine facilities of any one of these. When a node temporarily leaves a logical channel to which a node has already joined, the middleware program of the node searches its logical channel owner management table and indicates a temporary leave request to the searched logical channel owner. When data is unicasted and data indicating temporary leave permission is received from the logical channel owner, the logical channel owner node in the logical channel owner management table managed by itself is treated as a paused state,
The logical channel owner, when issuing the leave permission, treats the node that has transmitted the data indicating the temporary leave request from the logical channel / subscriber management table managed by itself as a suspended state. Item 5. The communication system between machine facilities according to any one of Items 1 to 4. When a machine device leaves the network itself, the logical channel owner table is searched to detect whether or not there is a logical channel object whose own node is a logical channel owner. Data indicating logical channel / owner substitution request is transmitted to all Subscriber nodes registered in the logical channel / subscriber management table, and the channel owner conformity is transmitted from the node corresponding to the logical channel / owner substitution request. Receiving the sex information, selecting one of the Subscriber nodes determined to be optimal, and transmitting all the logical channel / Subscriber management table data, executing the logical channel / owner transfer process,
6. The logical channel / owner management table is searched, a logical channel participating as a Subscriber node is detected, and the withdrawal from the logical channel or the temporary withdrawal process is executed. Communication system between machine facilities. In two or more machine facilities, a network resource that is a data communication means between the machine facilities, an application program, and a middleware program that enables machine-to-machine asynchronous multicasting, and a logic constructed between the machine facilities A communication method between machine facilities in which the machine facilities belonging to a channel transmit and receive data,
All of the middleware programs comprise a logical channel owner management table that stores data relating to logical channel owners that exist for each logical channel,
In the middleware program that is the logical channel owner, a logical channel that is a node belonging to the logical channel, a logical channel that stores data related to the Subscriber, a Subscriber management table,
The middleware program that has transmission data and is a logical channel Publisher executes logical channel data transmission / reception processing for searching the logical channel owner management table and transmitting the transmission data to the logical channel owner node,
The middleware program, which is the logical channel owner, executes a logical channel data collection and distribution process for distributing the transmission data to a registration destination node using the logical channel / subscriber management table. Machine-to-machine communication method characterized in that type asynchronous multicast communication is possible. In two or more machine facilities, a network resource that is a data communication means between the machine facilities, an application program, and a middleware program that enables machine-to-machine asynchronous multicasting, and a logic constructed between the machine facilities The machine equipment communication program for transmitting and receiving data between the machine equipment belonging to the channel,
All of the middleware programs comprise a logical channel owner management table that stores data relating to logical channel owners that exist for each logical channel,
In the middleware program that is the logical channel owner, a logical channel that is a node belonging to the logical channel, a logical channel that stores data related to the Subscriber, a Subscriber management table,
The middleware program that has transmission data and is a logical channel Publisher executes logical channel data transmission / reception processing for searching the logical channel owner management table and transmitting the transmission data to the logical channel owner node,
The middleware program, which is the logical channel owner, executes a logical channel data collection and distribution process for distributing the transmission data to a registration destination node using the logical channel / subscriber management table. Type machine-to-machine communication program characterized by enabling asynchronous multicast communication.

Images