Solved Example | Graphical techniques of inference | Fuzzy logic
HTML-код
- Опубликовано: 3 фев 2021
- Topics Covered:
03:18 - Mamdani method
08:45 - Sugeno method
11:45 - Tsukamoto method
Links:
Mamdani Systems - • Mamdani Systems | Grap...
Sugeno Systems - • Sugeno Systems | Graph...
Tsukamoto Systems - • Tsukamoto Systems | Gr...
Mam, I can not express happiness due to the clarity that I got by watching the video. Thank you for your support.
So glad to hear that:)
Excellent.
Many thanks! :)
Your video of Tsukomoto method is private. I need that video plz
It's Good lectures and there is no T
sukamoto method playlist or lecture 18 can you provide the link
Hello, thank you for your valuable video. Could you explain for me about the implication step and aggregation step in fuzzy? I saw in Fuzzy Logic Toolbox in Matlab, there are 2 methods for implication (min, prod) and 3 methods for aggregation (max, sum, probor). I do not know how the system will process with those options. Thank you for your support
In mamdani method, we have studied two cases of implication- max min and max product systems. In max min systems, we always obtained a truncated membership function as the output. So if you use the min implication method in matlab fuzzy logic toolbox, your final output fuzzy set will have a truncated membership function. Similarly the prod implication method in fuzzy logic toolbox corresponds to the max product method taught in the fuzzy logic lecture. That is, you will receive an output fuzzy set with scaled membership function. So min will give you a truncated membership function and prod will give you a scaled membership function. Please refer lecture on Mamdani Systems for detailed explanation on max min and max product implication methods - ruclips.net/video/fqZvzQayx7A/видео.html
For aggregation, the output fuzzy set of each rule will be aggregated or combined into one fuzzy set. The input of the aggregation process will be output truncated fuzzy sets obtained from different rules and the output of the aggregation process will be one fuzzy set combining all the output truncated fuzzy set. Three types of output truncated fuzzy sets are supported by the aggregation process- max(maximum), probor(probabilistic OR). nd sum ( sum of the rule output sets). For more information on aggregation refer to this link: www.mathworks.com/help/fuzzy/fuzzy-inference-process.html
@@Topperly Thank you so much for your support. I have already understood about the implication cases. However, in aggregation, could you explain for me the differences between 3 types of output truncated fuzzy sets: max, probor and sum? I tried to build a fuzzy system and tried those 3 different methods, but I did not see the difference
Please what's the difference between Mamdai , sugeno, Tsukamoto method?
Hi Inas,
They are different inference methods and are not related.
Please refer lectures 17, 18 and 19 of our Fuzzy Logic videos for details :)
Please upload another video on concepts of fuzzy logic
We have a playlist on Fuzzy logic which covers most of the topics. Please check it out! :)
Link : ruclips.net/p/PLhdVEDm7SZ-Ph7E3bYW89UbjD6zkW-vbf
Mam In diagrams how to draw a membership functions triangle , can you please explain
Hi Durga,
In real life, membership functions are created with help of experts who have experience in the field and by curve fitting with available data points :)
Hi, if i used sugeno method how i convert it to arduino code?can u help me please🙏 by what
Hi Noor,
I'm not familiar with Arduino platform, but I'm sure MATLAB has toolboxes for this method. You can check the logic behind MATLAB toolboxes and implement the same in your Arduino code :)
@@Topperly thank u so much i will do🙏
please send me the tsukamoto method video link
ruclips.net/video/hnWhElpYSso/видео.html
All three methods gave different z* value. How will we decide which one is correct or closest ?
Hi Kritika,
There's no value of z* which is correct or closest. Sugeno, Mamdani and Tsukamoto are independently developed methods which compute z* in their own way. We have to choose a method which best suits our application :)
Thanks for the reply. Do you have a video on detailed explanation of membership function like s shaped, gaussian etc. As in how we can choose which membership function, which one will work best in what scenario? That would be extremely helpful. Thanks again.