Design of a TRFN Controller for a Nonlinear Multivariable Process
B. Subathra, Nidhishree R Rao, Thota Karunakaran Radhakrishnan · Instrumentation Science & Technology · 2009
Feedback architecture is found to be more suitable for controlling dynamic processes, as it utilizes the past information of the process variables as against a controller designed using a feedforward structure that utilizes only the present information about the process variables. A comparative analysis of having a recurrent fuzzy controller over a conventional fuzzy controller in a control loop involving a non-linear dynamic system is enunciated using a laboratory two-tank heating process. A considerable improvement in performance is obtained by employing the recurrent structure in place of the conventional fuzzy controller. In addition, this proposed control strategy is numerically more efficient than the feedforward fuzzy structure. Results illustrate the effectiveness of the proposed structure.