Do my fuzzy logic correct and how to tune my PID properly using this system?

This is my system for first transfer function

This is the second one both of my system does not achieve the desired result and the second transfer function are worse

this is the mf of the system

i include my simulink here hope anybody can help my Final year project

[Sam Chak]

Hi @Isyraf Izuddin

Could you list out the desired performance requirements?

• percentage overshoot
• settling time
• steady-state error

[Isyraf Izuddin]

there no specific desired performance but the smaller the %os, faster settling time, and smallest steady state, faster rise time the better

[Sam Chak]

The 2nd transfer function is more difficult to control.

What are the logical/scientific reasons for choosing 7 equally-spaced centers of the membership functions? It helps to understand the behavior of fuzzy control rules.

More importantly, can you show the control surface using the gensurf command? Please post the image of the surface here.

[Isyraf Izuddin]

about the mf the parameter is [-3 3] for the inputs sorry i'm not realise it.

about the What are the logical/scientific reasons for choosing 7 equally-spaced centers of the membership functions? It helps to understand the behavior of fuzzy control rules.

is this help?

[Isyraf Izuddin]

this is the right parameter that i use

[Isyraf Izuddin]

Alright my supervisor have finalised to use the first transfer function

how do i improve my response to achieve the 1 with no %os, fast rise time, good settling time

this is the new mf and rule that i use

Answer 1

[Sam Chak]

Hi @Isyraf Izuddin

Plant #2

Since you didn't specify the performance criteria, the fuzzy system is designed with some improments as compared to the open-loop response of the original plant. In fact, it is very difficult to design a "pure" fuzzy system from the expert knowledge approach to satisfy the performance criteria. From the numerator polynomial and the characteristic polynomial, one can guess that the design process is not easy.

A SISO Sugeno Type-1 fuzzy system is designed. In the closed-loop system, the settling time is achieved under 300 s without overshoot. The original plant settles at 1,500 s.

Three Gaussian fuzzy sets are created for the fuzzy input with MF1 is gaussMF(Error, 0.03317, -0.146), MF2 is gaussMF(Error, 0.3098, 0.436), and MF3 is gaussMF(Error, 0.3286, 0.9175).

Three singleton fuzzy sets are created for the fuzzy output such that SGT1 is defined exactly as –144, SGT2 is defined exactly as 0.07204, and SGT3 is defined exactly as -0.009993. The Fuzzy Rules are given by

Rule 1: If Error is MF1, then Control Output is SG1.

Rule 2: If Error is MF2, then Control Output is SG2.

Rule 3: If Error is MF3, then Control Output is SG3.