Hi Erika Carmen,

If you already have the 'traditional' control method, then there is no need to design the fuzzy controller from scratch. The idea is to fuzzify the existing mathematical-based controller and turn it into a fuzzy controller.

Some new learners have trouble in the fuzzy control design because they have get used to the mathematical-based controller. Imagine that you always use a calculator to solve the arithmetic problems. If the calculator is suddenly taken away, you definitely don't feel comfortable in performing calculations in your head.

Similarly, you don't feel comfortable to construct the If-Then rules in the Fuzzy System. In fact, you probably have no idea how and where to begin. You maybe understand that IF the Error is Large, THEN the Control Action must Large too (in order to bring down the Error). But you don't know how Large they should be. This is the core problem.

In the mathematical-based controller, you would instantly know the magnitude of the Control Action by applying the Error Feedback Control Law (a mathematical formula). You probably just need to tune a few control gains (K1, K2, K3) to obtain the desired performance.

Naturally, the advice is to break out of your "control design" comfort zone.

Hi Erika Carmen,

If your 'traditional' control method is a standard linear Proportional plus Integral (PI) controller, then a 2-input triangular MF-based fuzzy system with 4 basic rules is sufficient to do the job. Each input (P and I) should have 2 MFs.

It is unnecessary to complicate the fuzzy system by having 9 rules or more.

Hello Yew-Chung Chak

Thanks so much for your interest on my questions.

I have a model which use incremental conduction techniques + regulator techniques and another one with pertub and observe.

Hi Erika Carmen,

As long as the incremental conduction techniques, regulator techniques, and the P&O can be expressed in the forms of non-stiff mathematical equations, they can be 100% fuzzified.

I do not know exactly inside them but it may be useful for your comparison. you can take a look for them.

https://www.mathworks.com/matlabcentral/fileexchange/64414-fuzzy-based-grid-integration-of-photovoltaic-power-system-using-multilevel-inverter?s_tid=prof_contriblnk

https://www.mathworks.com/matlabcentral/fileexchange/63195-fuzzy-logic-controller-based-mppt-of-hybrid-pv-and-wind-microgrid-system-with-bms

You can look my publishing papers about fuzzy logic controller in matlab simulink...