Simplify fuzzy control implementation
Kenneth E. Stoll, Patricia A. S. Ralston, S. Ramaganesan · Hydrocarbon processing · 1993
A controller that uses fuzzy rules provides better response than a conventional linear controller in some applications. The rules are best implemented as a breakpoint function. A level control example illustrates the technique and advantages over proportional-integral (PI) control. In numerous papers on fuzzy controller development, emphasis has been primarily on formal inferencing, membership functions, and steps in building a fuzzy relation, as described by Zadeh. The rationale used in formulating the required set of rules is usually neglected, and the interpretation of the final controller as an input-output algorithm is overlooked. Also, the details of fuzzy mathematics are unfamiliar to many engineers and the implementation appears cumbersome to most. Process description and control instrumentation. This article compares a fuzzy controller designed by specifying a breakpoint function with a traditional PI controller for a level control system on a laboratory scale. In this discussion, only setpoint changes are considered.