Abstract:
Sensitivity and Robustness is the primary issue while designing the controller for non-linear systems. One of the performance objectives for controller design is to keep the error between the controlled output and the set-point as small as possible. Controlling the speed and position of a DC Motor in all the condition is not possible by conventional control technique. The classical control methods have its limitations. Most of these techniques are based on hit and trial method. The response of the system is also not proper. And in the case of disturbance or parameter variation it fails to control the system.
PID controller is one of the most basic and successful controller. It controls the disturbance and parameter variation up to some extent. But the response of the system changes to a large extent. For controlling the response Fuzzy Controller is used. The fuzzy controller helps in controlling the system non-linearity.
A much improved system response can be achieved by Self Tuned Fuzzy Controller, which adjust the gain of controller according to the system variation or disturbance. However this response can be further improved by the systematic analysis of system response. Manually we can adjust the fuzzy rules by analysing the effect of each rule over the system. This analysis can be made through the graphical analysis of the system response.
Adaptive Fuzzy Controller is the most advanced form of controller. It adjusts the fuzzy logic based controller to give better response. We have used a Tabular Based Adaptive Fuzzy Controller to control the system non-linearity. It is a sugeno type controller, which adjust the rule base according to the system variation.
Controlling the position of a DC Motor is a major control issue. A Neuro-Fuzzy Controller helps us to design a controller which is build from the data base of the system. The FIS generated after training is very effective in controlling the position of DC Motor. As it is compared with the PID controller under loaded condition, the response of the controller generated through ANFIS is more robust than PID. This shows that FIS generated have more adaptability than normal PID controller.
