Fuzzy Control Surface

Temperature 24 °C

Humidity 55 %

Fan speed (crisp)

–

Rules firing

–

Dominant rule

–

How it works

Where the 2D version plots membership curves one input at a time, this scene renders the controller's entire response as one continuous surface: for every (Temperature, Humidity) pair on the base grid, the same Mamdani pipeline — fuzzify, apply the 3×3 rule base with min-AND, clip and aggregate the output sets, defuzzify by centroid — is evaluated once to get a height, then the whole surface is meshed and colour-mapped by that height.

height(T,H) = centroid( agg( min(strength, μ_out) ) )
strength_rule = min(μ_T(T), μ_H(H))

Control surface — the mesh itself; ridges and plateaus reveal where the rule base is smooth (triangular memberships) versus abrupt (shoulder edges).
White marker — sits exactly on the surface at your current Temperature/Humidity input, at the height equal to the crisp output.
Rule pillars — nine translucent columns on the base plane, one per rule cell, whose height is that rule's firing strength (min of the two current memberships) for the current input — the 3D analogue of the 2D rule-grid heat map.
Auto-orbit — slowly rotates the camera so the surface's shape reads clearly without manual dragging.

A genuine 3D control surface like this is exactly what a fuzzy-controller designer inspects to check the response is smooth and monotonic before deploying the rule base to hardware.

Low fan speed Medium fan speed High fan speed Current input

⚠ Couldn't load the 3D engineThree.js failed to load from the CDN. Check your connection and reload.

Fuzzy Control Surface — 3D Simulator