PHD Discussions Logo

Ask, Learn and Accelerate in your PhD Research

Question Icon Post Your Answer

Question Icon

How can a gray number be converted to a fuzzy number?

In my systems modeling work, I encounter gray numbers representing uncertain intervals. To use fuzzy logic controllers, I need to convert these into proper fuzzy numbers with membership functions. I’m unsure about the principled approach whether it’s a direct mapping or if it requires additional assumptions about the data distribution within the gray interval.

 

All Answers (1 Answers In All)

By Adi Answered 11 months ago

In my applied work with uncertain engineering parameters, this conversion is about interpreting the gray interval’s "potential." A gray number gives you a range, but no distribution. To convert it, you must impose an assumption. I typically recommend a triangular fuzzy number: the gray interval’s bounds become the support (membership = 0), and you choose a most plausible point, often the midpoint, for the peak (membership = 1). This is pragmatic. For more nuance, if you have historical data hinting at a distribution within the interval, you can shape a trapezoidal or Gaussian-like fuzzy set to reflect it.

Your Answer