Time-Delay Systems

Abstract

A time-delay under the context of process control systems may be defined as the time interval from the application of a control signal to any observable change in the process variable. Time delays have always been among the most difficult problems encountered in process control. It could occur for various reasons and in different magnitudes. It may be attributed to a delay in measurement of the process variable. One such example is the steel rolling process where the measurement point is located at some distance downstream of the steel press so that there is always a time delay in the feedback of plate thickness to the press controller. The time delay could be due to a delay in control, typical of processes involving transportation of materials in pipelines and conveyor systems such as thermal, composition and distillation processes. For these processes, the actuating devices are physically limited, so that instantaneous impulse-like action is not realizable. In modern digital control systems, time-delay can also arise in the form of sampling intervals, or the inherent polling and waiting states involved in collision avoidance for multiple access field-network systems. The best approach to eliminate time-delay is to avoid it altogether, through careful and proper system and industrial engineering practices. In the design of processes, careful considerations must be given to placement of sensors and the control devices to minimize dead times. However, under physical, spatial and operational constraints in the practical implementation of these systems, the extent to which this is realizable is fairly limited. Therefore, control of time-delay systems represents one important class of problems which has and will continue to pose a strong challenge for process control engineers.