Fleet & map

Delivery stops: 22

Annealing schedule

Cooling rate α: 0.9950 Initial temperature: 60 Steps per frame: 10

Run controls

Live stats

Iteration0

Temperature—

Current distance—

Best distance—

Acceptance rate—

Improvement vs start—

Each step proposes a random 2-opt move (reverse a route segment). Moves that shorten the tour are always accepted; moves that lengthen it are accepted with probability e^(−Δ/T) — the Metropolis criterion. As T cools geometrically (T ← α·T), the tour settles toward a near-optimal loop while early high-T jumps let it escape local minima.