23.03.2017 Views

wilamowski-b-m-irwin-j-d-industrial-communication-systems-2011

You also want an ePaper? Increase the reach of your titles

YUMPU automatically turns print PDFs into web optimized ePapers that Google loves.

Communication Aspects of IEC 61499 Architecture 55-19<br />

This investigation shows that CIP fits very well to the execution and object model of IEC 61499.<br />

Our example made the assumption that all devices on the CIP network are IEC 61499-enabled<br />

control devices. However, several advantages can be gained from combining legacy CIP devices<br />

with IEC 61499-enabled devices. Two different application scenarios should be considered: first IEC<br />

61499-enabled CIP slave device (i.e., programmable I/O devices) and second use standard CIP slave<br />

devices from IEC 61499 devices. In the first case, the CIP interface of the IEC 61499 device would<br />

be specified by utilizing the CIP FBs, and with this mechanism the interface can be adapted to the<br />

specific IEC 61499 application running in the device. On the other end of the <strong>communication</strong>, a CIP<br />

master device (e.g., a PLC) would interact with the IEC 61499 device in the same way as with any other<br />

CIP slave device. However, the use of IEC 61499 allows to pre-process I/O data and to provide higher<br />

level information to the PLC. This use case is an important step toward a fully distributed control<br />

system as described by IEC 61499. In the second use case, the IEC 61499 control device will take the<br />

place of the PLC, and the CIP <strong>communication</strong> FBs will mimic the behavior of a CIP master. However,<br />

in contrast to the existing PLC interface where the network data are presented to the application as a<br />

flat memory region, IEC 61499 allows a structured access to the data. This is achieved by representing<br />

each slave device as a separate FB. This allows to leverage existing investments in plants and I/O device<br />

development, or to extend dumb I/O modules by adding own CPUs. Therefore this will be an important<br />

part of an IEC 61499-based control environment.<br />

Both use cases are not specific to the CIP environment. They can be applied to any fieldbus system<br />

used in <strong>industrial</strong> control <strong>systems</strong>.<br />

55.11 Impact of Communication Semantics<br />

on Application Behavior<br />

One word of caution however. Traditional control applications running on a single controller periodically<br />

or cyclically read from, and write to, all the remote devices connected to the network bus (even<br />

when no changes to the state of these devices is required); so if the network suffers a transient error during<br />

one period/cycle, the application will naturally recover in the next period/cycle, when a new network<br />

message is generated. In contrast, an IEC 61499 application is event oriented, so it will typically only<br />

generate a network message when the PUBLISH or SEND FBs receive an REQ event, or when an RCV<br />

FB receives an RSP event. While this may be considered an advantage due to the lower traffic imposed<br />

on the network, it also has the drawback that the application becomes much more susceptible to transient<br />

<strong>communication</strong> errors that may occur on this same network.<br />

Consider, for example, the system in Figure 55.13. Now imagine that a CR event generated by FB Logic1<br />

does not reach the NEW event input of Logic2 due to a transient <strong>communication</strong> error on the network<br />

used by the publish and subscribe CR2 FBs. The result will be that all the lights connected to Logic1 will<br />

be correctly switched off, and no light connected to Logic2 will be lit to continue the chasing sequence.<br />

We may conclude that due to the asynchronous event-based architecture used by IEC 61499 applications,<br />

these are more susceptible to network failures. Nevertheless, this does not mean that the application<br />

will fail at the smallest network interference. In fact, many <strong>communication</strong> protocols include error<br />

detection and correction algorithms. If the <strong>communication</strong> protocol used between the CR2 PUBLISH<br />

and SUBSCRIBE FBs includes error detection and correction mechanisms, the IEC 61499 application<br />

will not be affected by the transient network failure. However, note that in the presence of failures, the<br />

exact semantics of the <strong>communication</strong> service interface FBs is not specified by the IEC 61499 standard,<br />

as well as the choice of which specific implementation of the <strong>communication</strong> service interface FBs will<br />

affect how the IEC 61499 application will react when in the presence of transient network failures.<br />

The message delivery semantics in the presence of failures may generally be classified into three<br />

classes: “at most once,” “at least once,” and “exactly once.” Although the “exactly once” semantics are<br />

usually the preferred semantics, due to the higher number of messages required, sometimes semantics<br />

© <strong>2011</strong> by Taylor and Francis Group, LLC

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!