Fuzzy logic is a relatively new control technology. In the mid-1970s Mamdani began to investigate fuzzy logic as an alternative to the PID controller and sparked the first practical applications of fuzzy logic in the process control industry. The controller followed rules that generated an output by evaluating the error and the change in error. Membership functions that represent the degree of truth as an extension of valuation were created for error, change-in-error, and change-in- output. The shape, size, and overlap of membership functions were defined to be tunable parameters. Similar to the adjustment of PID proportional, integral, and derivative gain, more or less aggressive control for a given process can be achieved with fuzzy logic control through adjusting the scaling of membership functions.
As addressed in Chapter 6 of Advanced Control Foundation – Tools, Techniques, and Applications, the design and implementation of fuzzy logic control to replace PID control was simplified Professor Joe Qin while he worked at Emerson. He proposed methods for basing the adjustment of membership function scaling on the identified process gain and dynamics. Today, based on these developments, many single loop digital controllers, and some distributed control systems, offer a version of fuzzy logic control with an auto-tuning capability.
The control functionality that may be achieved using fuzzy rules and membership functions is similar to that available for PID control. For some specific process applications, fuzzy logic control enables faster setpoint recovery with less overshoot than PID control for both setpoint and load changes as illustrated below.
This figure compares the Integral of Absolute Error (IAE) on two cascade loops, one using traditional PID and the other using fuzzy logic control. Fuzzy exhibited less IAE for both load disturbance and setpoint changes. The difference in control response is illustrated below.
Fuzzy logic is best suited for controlling processes that are characterized by large time constantsand little or no deadtime and where the control objective is to reach setpoint in the minimum time with no overshoot. In the next few blog I will be addressing different aspects of fuzzy logic control.