Conditions for the t-norm extended mapping to preserve the convexity of fuzzy sets (Q678962)

A map \(A: \mathbb{R}\to I:=[0,1]\) is called a convex fuzzy set if \(A(\alpha x+(1-\alpha)y)\geq\min(A(x),A(y))\) for all \(x, y\in\mathbb{R}\) and \(\alpha\in I\). Let \(T:I\times I\to I\) be a t-norm (that is, \((I,T)\) is an Abelian monoid with unit 1 and such that \(T(\alpha,\beta)\leq T(\gamma,\delta)\) whenever \(\alpha\leq\gamma\) and \(\beta\leq\delta\)). For \(f:\mathbb{R}^2\to\mathbb{R}\) let \(f_T: I^{\mathbb{R}}\times I^{\mathbb{R}}\to I^{\mathbb{R}}\) be defined by \(f_T(A,B)(z)= \sup\{T(A(x),B(y)): f(x,y)=z\}\) (note: \(\sup\emptyset=0\)). Main result: If \(f:\mathbb{R}^2\to\mathbb{R}\) is continuous and \(A,B:\mathbb{R}\to I\) are convex fuzzy sets, then so is \(f_T(A,B)\).