Abstract :
First examines the fundamental principles of distributed control. Basically, all transducers-sensors and actuators-have embedded intelligence and are therefore referred to as smart. They are interconnected by networks of various topologies which come under the generic name of fieldbuses. However it is useful to divide the fieldbuses into two categories, transducer networks which interconnect the smart transducers and strategy networks which interconnect control devices which we refer to as strategy engines. At the top level are the computing devices interconnected by computer networks-intranets and internets. This implies a three-level network hierarchy but in practise this can be collapsed or enlarged. The basic attractions of such a system are, first of all, it can potentially give very significant savings in material cost-principally wiring, connectors and cubicles-and in installation time: also it is capable of better overall control, easier maintenance and greater integrity. The principle difficulty in implementing such a system is that unless the system is open whereby a set of standards are agreed allowing a true multi-vendor system employing best of class principles, it can only be realised by a single vendor solution which can never be optimal and carries significant risk. This paper looks at various standardisation initiatives, both international and de-facto, focuses on the most significant, and attempts to see how an optimal solution might be achieved
