Hi @ali
To tune an initially-designed fuzzy controller using the app, you can generally follow the five simple steps outlined below:
Step 1: Read the data from the table in .xlsx or .csv format. The data sets will be loaded on the Workspace. In the Command Window, enter the following:
T = readtable("Data.xlsx", VariableNamingRule="preserve");
x = T{:,1}; % assign input data as 'x'
y = T{:,2}; % assign output data as 'y'
Then, open your designed fuzzy controller in the Fuzzy Logic Designer app. On the TUNING tab, load the input and output data (assigned as 'x' and 'y' in this example).
Step 2: Identify the desired tunable parameters for the input MFs, output MFs, and the If-Then rules. If you do not wish to tune certain parameters (typically known by experienced users and knowledge experts), uncheck the parameter index in the bottom right panel.
Step 3: Click the 'Tuning Option' button and select the desired optimization method for tuning purposes. The Genetic Algorithm is the default method.
Step 4: Click the 'Tune' button to initiate the tuning process. You may click the 'Stop' button when the desired accuracy is achieved.
Step 5: The final step is to visualize the error distribution. If the results are not satisfactory, adjust the tunable parameters, set constraints, and restart the tuning process.
