Simulation - number of samples

Question

Steve 2024-7-27

0
链接

此问题的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/2140666-simulation-number-of-samples

编辑： Torsten 2024-7-28

MATLAB Helpcenter states:

When analyzing your FIS, you can configure the resolution of the output variable universe of discourse. To do so, set the Number of Samples parameter.

Is there a preset (optimum) value and how is this determined?

0 个评论
显示 -2更早的评论隐藏 -2更早的评论

请先登录，再进行评论。

请先登录，再回答此问题。

Answer 1

Torsten 2024-7-27

1
链接

此回答的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/2140666-simulation-number-of-samples#answer_1491316

编辑：Torsten 2024-7-27

Is there a preset (optimum) value and how is this determined?

As big as necessary and as small as possible. You will have to test it for several values.

2 个评论
显示无隐藏无

Steve 2024-7-28

Can you please elaborate.

Torsten 2024-7-28

编辑：Torsten 2024-7-28

Do you know Monte-Carlo simulation ? You need a minimum number of runs to establish a trustworthy value for probability, but you shouldn't use too many runs in order not to waste time and memory.

I'm not sure, but I think "Number of Samples" here is a similar parameter as the number of runs in Monto-Carlo Simulation. The more you choose, the better will be the resolution of the result, but the more computing time and memory will be necessary.

From the documentation:

When analyzing your FIS, you can configure the resolution of the output variable universe of discourse. To do so, set the Number of Samples parameter.

But before using a tool like the fuzzy logic toolbox, you should read its documentation to know about the theoretical background. Or do you know better what is meant by "Mamdani To Sugeno" or "Rule Inference" or "Control Surface" ?

请先登录，再进行评论。